Combinations of compounds for developing in vitro models of the blood-brain barrier

Abstract

A combination of drugs useful for establishing or improving the blood-brain barrier (BBB) ​​in an in vitro model is provided. Specifically, the combination includes a cAMP / PKA signaling pathway activator, a cAMP-specific PDE inhibitor, a Wnt signaling pathway activator, and a TGF-β receptor inhibitor. Use of the drug combination in a cell culture or in an in vitro model of the BBB increases the tightness of the paracellular and / or transcellular barrier formed by cells in culture or in an in vitro model of the BBB.

Description

[Technical Field]

[0001] A combination of drugs useful for establishing or improving an in vitro model of the blood-brain barrier (BBB) ​​is provided. Specifically, the combination includes a cAMP / PKA signaling pathway activator, a cAMP-specific PDE inhibitor, a Wnt signaling pathway activator, and a TGF-β receptor inhibitor. Use of the drug combination in a cell culture or in an in vitro model of the BBB increases the tightness of the paracellular and / or transcellular barrier formed by cells in culture or in an in vitro model of the BBB. [Background technology]

[0002] The cerebrovascular endothelium, which forms the blood-brain barrier (BBB), possesses unique properties that protect the brain and maintain homeostasis. Microvascular endothelial cell (EC) characteristics are organ-specific, and signals from brain cells are crucial for inducing the physical, metabolic, and chemical barrier properties that contribute to the brain-specific phenotype of ECs. Maturation of intercellular junctions, downregulation of nonspecific vesicular transcytosis, and increased expression of specific transporters, receptors, and enzymes are key characteristics of BBB development. However, these characteristics are not inherent to vascular ECs; rather, they are induced and maintained by the local microenvironment of central nervous system (CNS) capillaries.

[0003] Endothelial cells (ECs) lining the smallest blood vessels in the brain possess unique anatomical and functional properties collectively known as the BBB (Zlokovic 2008). Through specialized tight junction complexes, low-frequency vesicular transcytosis, a negatively charged glycocalyx, and efflux and influx transport systems, the BBB precisely controls the composition of the neural microenvironment (Abbott 2010). While this protects the brain from harmful substances, the barrier also prevents most drugs from entering the central nervous system, hindering drug development for neurological disorders (Pardridge 2012). Furthermore, BBB disruption is a hallmark of a range of neuropathological disorders (Munji 2019, Sweeney 2019), yet these processes are incompletely understood at the molecular level in humans. In vitro models that faithfully mimic the human BBB are needed to improve prediction of drug delivery to the brain and provide mechanistic insights into BBB function (and dysfunction).

[0004] The current gold-standard culture model of the BBB is based on primary brain endothelial cells isolated from animal tissues (Helms et al. 2016, Veszelka et al. 2018). However, interspecies differences in the levels of BBB transporters and receptors exist (Uchida et al. 2020), making the development and characterization of reliable human cell-based BBB models crucial. Over the past decade, BBB models based on human umbilical cord blood stem cells and induced pluripotent stem cells (hiPSCs) have been established (Lippmann et al. 2012, Cecchelli et al. 2014). HiPSC models adapted in several laboratories exhibit a partial BBB-associated phenotype, with extremely high paracellular tightness and a morphology more typical of epithelial cells. Recent studies have revealed that hiPSC-derived "one-stage" BBB models exhibit neuroectodermal epithelial, rather than endothelial, characteristics due to inappropriate selection of differentiation factors (Lu et al. 2021).

[0005] Other models use a "two-step" approach: First, vascular ECs are generated from iPSCs (Nishihara et al. 2020) or CD34-positive umbilical cord blood stem cells, then differentiated into a brain-like phenotype through selective culture conditions and coculture with brain microvascular pericytes (Cecchelli et al. 2014). These models express endothelial properties, but the tightness of the paracellular barrier is weaker than that of primary cell-based models (Helms et al. 2016, Veszelka et al. 2018) (Cecchelli et al. 2014, Nishihara et al. 2020).

[0006] The gold standard culture model of the BBB uses primary cerebral microvascular ECs isolated from animal tissue (Helms 2016). However, there are significant interspecies differences in the BBB proteome related to drug transport (Uchida 2011, Shawahna 2011, Uchida 2020), which negatively impacts the translatability of findings from animal models to clinical trials. However, there are significant interspecies differences in the BBB proteome related to drug transport (Uchida 2011, Shawahna 2011, Uchida 2020), which may negatively impact the translatability of findings from animal models to clinical trials. Several BBB models have been established using induced pluripotent stem cells (Lippmann 2012, Lippmann 2014, Canfield 2017, Appelt-Menzel 2017, Praca 2019, Stebbins 2019, Qian 2019, Vatine 2019, Park 2019, Linville 2019, Nishihara 2020) or umbilical cord blood-derived CD34-positive hematopoietic stem cells (Cecchelli 2014, Missu 2019, Moya 2021) as human surrogates with excellent scalability. While the development of these protocols represents a major technological goal, the resulting models often exhibit a mixed epithelial-endothelial phenotype or have suboptimal barrier properties that significantly limit their usefulness (Lippmann 2020, Lu 2021, PNAS). To address this, two-step differentiation has emerged as a promising strategy, where stem cells are first differentiated into ECs and then induced to develop brain-like features by various techniques, including co-culture with other brain cell types (Cecchelli 2014, Praca 2019, Nishihara 2020).

[0007] BBB properties are not inherent to ECs; rather, they are promoted and maintained in vivo by organ-specific signaling cues from pericytes (Armulik 2010, Daneman 2010, Sweeney 2016) and astrocytes (Janzer & Raff 1987, Simard 2003, Abbott 2006). However, selecting the correct factors to enhance barrier integrity in addition to in vitro coculture signals remains challenging. Treatment with cAMP-elevating agents (cPT-cAMP, forskolin, and the cAMP-specific phosphodiesterase inhibitor Ro20-1724) has previously been shown to increase resistance across EC monolayers, both alone and especially in combination with astrocyte-conditioned medium (Rubin 1991, Wolburg 1994, Deli 1995). During development, Wnt / β-catenin signaling regulates cerebral angiogenesis (Stenman 2008, Daneman 2009, Vanhollebke 2015) and BBB formation (Liebner 2008, Cullen 2011, Zhou 2014), and this pathway can be activated in vitro by Wnt ligands (Wnt-3a, Wnt-7a / b) or small molecules (LiCl, 6-BIO, CHIR99021) to induce a subset of barrier properties (Paolinelli 2013, Weksler 2013, Cecchelli 2014, Laksitorini 2019, Gastfriend 2021). Recently, TGF-β receptor antagonists (RepSox, A83-01, but not SB431542) have also been demonstrated to increase junctional tightness in stem cell-derived ECs (Roudnicky 2020, PNAS, Yamashita 2020) via claudin-5, a major tight junction protein in the BBB (Nitta 2003, Greene 2019). Summary of the Invention

[0008] In a first aspect, the present invention provides the use of a combination of compounds in a culture of cells, the combination comprising: cAMP / PKA signaling pathway activators, a cAMP-specific PDE inhibitor; a Wnt signaling pathway activator; and a TGF-β receptor inhibitor, comprising, consisting essentially of, or consisting of.

[0009] Preferably, the cell culture is, or is part of, or is present in an in vitro model of the blood-brain barrier (BBB).

[0010] In a preferred embodiment, the use of the combination of compounds is for developing an in vitro model of the BBB (eg, for obtaining or improving an in vitro model of the BBB).

[0011] In a second aspect, the present invention provides a method for producing a medicament for the treatment of a pulmonary arthritis, comprising: cAMP / PKA signaling pathway activators, a cAMP-specific PDE inhibitor; a Wnt signaling pathway activator; The present invention relates to a cell culture to which a compound combination comprising, consisting essentially of, or consisting of a TGF-β receptor inhibitor and a TGF-β receptor inhibitor is applied or to which a compound combination is applied, and which is part of or present in an in vitro model of BBB. Preferably, the cell culture to which the compound combination is applied or to which the compound combination is applied is a BBB model. The use of the compound combination in the cell culture improves the characteristics of the BBB model compared to a model to which the compound combination is not applied. Thus, the present invention also provides an improved BBB model. Preferably, the improved BBB model is a monolayer culture of cells, including brain endothelial cells, treated with a combination of compounds; It has improved characteristics as defined in the Summary of the Invention or as defined herein, in particular in the Detailed Description of the Invention before the Experimental Results or as defined in the Examples, compared to a BBB model not treated with the compound combination.

[0012] According to a further aspect, the present invention provides an improved BBB barrier model, comprising: cAMP / PKA signaling pathway activators, a cAMP-specific PDE inhibitor; a Wnt signaling pathway activator; The present invention relates to a model comprising a monolayer culture of cells, including brain endothelial cells, treated with a combination of compounds comprising, consisting essentially of, or consisting of a TGF-β receptor inhibitor and a TGF-β receptor inhibitor, which comprises improved features as defined in the Summary of the Invention section or as defined herein, in particular as defined in the Detailed Description of the Invention before the Experimental Results or as defined in the Examples, compared to a BBB model not treated with said combination of compounds.

[0013] In a further aspect, the present invention provides a medium for a BBB model or BBB models based on the cells defined herein, in particular as defined in any paragraph of the Summary of the Invention, comprising: cAMP / PKA signaling pathway activators, a cAMP-specific PDE inhibitor; a Wnt signaling pathway activator; and a TGF-β receptor inhibitor.

[0014] In a preferred embodiment, the medium is one that is used to develop an in vitro model of the BBB (eg, to obtain or improve an in vitro model of the BBB).

[0015] According to a further aspect, there is provided a method for developing, creating or improving an in vitro model of the BBB, comprising contacting a combination of compounds with cells in culture or in an in vitro model of the BBB, The combination of compounds is cAMP / PKA signaling pathway activators, a cAMP-specific PDE inhibitor; a Wnt signaling pathway activator; and a TGF-β receptor inhibitor.

[0016] In a further aspect, the present invention provides a method for developing an in vitro model of the BBB (e.g., for obtaining or improving an in vitro model of the BBB), comprising: preparing a layer of a culture of cells on a porous support; cAMP / PKA signaling pathway activators, a cAMP-specific PDE inhibitor; a Wnt signaling pathway activator; and culturing the cells in a medium containing a combination of compounds comprising, consisting essentially of, or consisting of a TGF-β receptor inhibitor.

[0017] In a further aspect, the present invention also provides a kit for developing a model of the BBB, comprising: A culture of cells according to any paragraph of the Summary of the Invention, preferably comprising brain endothelial cells; cAMP / PKA signaling pathway activators, a cAMP-specific PDE inhibitor; a Wnt signaling pathway activator; a TGF-β receptor inhibitor; and a compound combination comprising, consisting essentially of, or consisting of, a TGF-β receptor inhibitor. and optionally an apparatus for preparing a cell culture layer, preferably comprising a porous substrate, preferably a collagen-coated porous substrate. Brain endothelial cell cultures, and / or Brain endothelial cell and astrocyte cultures, and / or Brain endothelial cell and pericyte cultures, and / or Contains cultures of brain endothelial cells, pericytes and astrocytes.

[0018] Preferably, the combination of compounds is formulated in the form of a medium.

[0019] In any one of the aspects of the present invention, Preferably, cells in culture or in an in vitro model of the BBB are contacted with the combination of compounds for at least 5 hours, preferably at least 10 hours, preferably at least 24 hours, preferably at least 36 hours, preferably at least 48 hours, preferably at least 72 hours.

[0020] Preferably, the cells in culture or in an in vitro model of the BBB are contacted with the combination of compounds for no longer than 72 hours.

[0021] Preferably, the cAMP / PKA signaling pathway activator is a cell-permeable, low molecular weight activator of the cAMP / PKA signaling pathway.

[0022] Preferably, the cAMP / PKA signaling pathway activator is a cell-permeable cAMP derivative (cAMP derivative) that has the effect of activating the cAMP / PKA signaling pathway, such as pCPT-cAMP, including variants such as dibutyryl-cAMP, 8-Br-cAMP, or 8-pCPT-2-O-Me-cAMP-AM.

[0023] In a preferred embodiment, the cAMP / PKA signaling pathway activator is forskolin or other direct PKA activators such as bucladesine, virovetin, or jaspermycin.

[0024] Highly preferably, the cAMP / PKA signalling pathway activator is cAMP or pCPT-cAMP.

[0025] Preferably, the cAMP-specific PDE inhibitor is a cell-permeable low molecular weight compound.

[0026] Preferably, the cAMP-specific PDE inhibitor is a competitive cAMP-specific PDE inhibitor.

[0027] Preferably, the cAMP-specific PDE inhibitor is a cAMP-specific PDE4, PDE7 or PDE8 inhibitor, more preferably a competitive cAMP-specific PDE4, PDE7 or PDE8 inhibitor, more preferably a cAMP-specific PDE4 inhibitor, more preferably a competitive cAMP-specific PDE4 inhibitor, preferably apremilast, crisaborole or roflumilast, very preferably Ro20-1724 (4-(3-butoxy-4-methoxybenzyl)imidazolidin-2-one).

[0028] Preferably, the Wnt signalling pathway activator is a cell-permeable low molecular weight compound.

[0029] Preferably, the Wnt signalling pathway activator is a Wnt downstream activator.

[0030] Preferably, the Wnt signaling pathway activator is a cell-permeable low molecular weight GSK-3β inhibitor, such as CHIR99021, 6-bromoindirubin-3-oxime (6-BIO); Or preferably a Li salt, a Li donor compound, a cell permeable low molecular weight Li donor compound such as Li orotate or Li carbonate, very preferably LiCl.

[0031] Preferably, the Wnt signaling pathway activator is Wnt-3a.

[0032] Preferably, the TGF-β receptor inhibitor is a cell-permeable low molecular weight compound (small molecule).

[0033] Preferably, the TGF-β receptor inhibitor is an ALK4 or ALK5 or ALK7 (ALK4 / 5 / 7) specific TGF-β receptor inhibitor, preferably an IC 50 These are high affinity inhibitors with values ​​below the micromolar range, preferably below 500 nM, more preferably below 300 nM or below 200 nM, and very preferably below 100 nM.

[0034] Highly preferably, the TGF-β receptor inhibitor is an ALK4 / 5 / 7-specific TGF-β receptor inhibitor, such as A-83-01 (3-(6-methyl-2-pyridinyl)-N-phenyl-4-(4-quinolinyl)-1H-pyrazole-1-carbothioamide).

[0035] Preferably, the TGF-β receptor inhibitor is an ALK4 / 5 / 7-specific TGF-β receptor inhibitor, such as bactosertib or SB431542.

[0036] In a preferred embodiment, the TGF-β receptor inhibitor is an ALK5-specific TGF-β receptor inhibitor, such as SB525334 or BIBF0775, very preferably RepSox.

[0037] In one embodiment, the TGF-β receptor inhibitor is A-83-01 (3-(6-methyl-2-pyridinyl)-N-phenyl-4-(4-quinolinyl)-1H-pyrazole-1-carbothioamide) or RepSox.

[0038] Preferably, the combination is cAMP, a cell-permeable low molecular weight competitive cAMP-specific PDE4 inhibitor, preferably apremilast, crisaborole or roflumilast, very preferably Ro20-1724, a cell-permeable low molecular weight Li donor compound, preferably a Li salt, very preferably LiCl, and an ALK4 / 5 / 7-specific TGF-β receptor inhibitor, very preferably A83-01; cAMP, a cell-permeable low molecular weight competitive cAMP-specific PDE4 inhibitor, preferably apremilast, crisaborole or roflumilast, very preferably Ro20-1724, a cell-permeable low molecular weight Li donor compound, preferably a Li salt, very preferably LiCl, and a cell-permeable low molecular weight ALK5-specific TGF-β receptor inhibitor, very preferably RepSox, cAMP, a cell-permeable low molecular weight competitive cAMP-specific PDE4 inhibitor, preferably apremilast, crisaborole or roflumilast, very preferably Ro20-1724, Wnt-3a and a cell-permeable low molecular weight ALK4 / 5 / 7-specific TGF-β receptor inhibitor, very preferably A83-01, or comprising, consisting essentially of, or consisting of cAMP, a cell-permeable low molecular weight competitive cAMP-specific PDE4 inhibitor, preferably apremilast, crisaborole or roflumilast, very preferably Ro20-1724, Wnt-3a and a cell-permeable low molecular weight ALK5-specific TGF-β receptor inhibitor, very preferably RepSox.

[0039] In a preferred embodiment, the cAMP is a cell-permeable cAMP analog, and in a highly preferred embodiment, pCPT-cAMP.

[0040] Preferably, the combination is The compound according to claim 1, further comprising: cAMP; a cell-permeable low molecular weight competitive cAMP-specific PDE4 inhibitor, preferably apremilast, crisaborole or roflumilast, very preferably Ro20-1724; a cell-permeable low molecular weight Li donor compound, preferably a Li salt, very preferably LiCl; and an ALK4 / 5 / 7-specific TGF-β receptor inhibitor, very preferably A83-01.

[0041] Preferably, the combination is The composition comprises, consists essentially of, or consists of a Wnt signaling pathway activator selected from cAMP, Ro20-1724, LiCl, Wnt-3a, and combinations thereof, and a TGF-β receptor inhibitor selected from A83-01, RepSox, and combinations thereof.

[0042] Preferably, the combination is a) cAMP, Ro20-1724, LiCl, and a TGF-β receptor inhibitor selected from A83-01, RepSox, and combinations thereof; b) comprising or consisting of cAMP, Ro20-1724, Wnt-3a, and a TGF-β receptor inhibitor selected from A83-01, RepSox, and combinations thereof.

[0043] In a highly preferred embodiment, a cell-permeable low molecular weight Li donor compound is used, preferably a Li salt, most preferably LiCl.

[0044] Particularly preferably, the combination is a) cAMP, Ro20-1724, LiCl and A83-01, and b) cAMP, Ro20-1724, Wnt-3a and A83-01; c) cAMP, Ro20-1724, LiCl and RepSox, or d) comprising, consisting essentially of, or consisting of a combination selected from the group consisting of cAMP, Ro20-1724, Wnt-3a and RepSox.

[0045] Highly preferably, the combination is a) cAMP, Ro20-1724, LiCl and A83-01, and b) comprising, consisting essentially of, or consisting of a combination selected from the group consisting of cAMP, Ro20-1724, Wnt-3a, and A83-01; Highly preferably, the combination consists of cAMP, Ro20-1724, LiCl and A83-01.

[0046] In a highly preferred embodiment, the cAMP is pCPT-cAMP.

[0047] Preferably, the cells in culture or in an in vitro model of the BBB or the cells of the improved BBB model or kit comprise cells selected from the group consisting of primary brain endothelial cells (ECs), brain-like ECs, glial cells, brain pericytes, cells from brain cell lines, preferably cultures of cells including vascular brain ECs derived from tissue (primary) or stem cells or induced pluripotent stem cells.

[0048] Preferably, the effect of the combination of compounds on barrier integrity, as represented by impedance, is detectable from cells in culture after 72 hours of treatment or from cells in an in vitro model of the BBB after treatment with the combination has ceased.

[0049] Preferably, the present invention relates to the use of a combination or method whereby the developed BBB model remains stable upon removal of the combination of small molecule compounds from the culture medium.

[0050] Preferably, the use or method is for obtaining / creating an in vitro model of the BBB.

[0051] In a preferred embodiment, the present invention relates to a use or method for developing an in vitro model of the BBB comprising endothelial cells (eg for obtaining or improving an in vitro model of the BBB).

[0052] Preferably, the use or method is for improving an in vitro model of the BBB.

[0053] Preferably, the use or method is for increasing the tightness of the barrier formed by cells in culture or cells in an in vitro model of the BBB.

[0054] Preferably, the use or method is for promoting maturation of the barrier formed by cells in culture or cells in an in vitro model of the BBB.

[0055] Preferably, the use or method is for enhancing brain endothelial cell identity in cells in culture or in an in vitro model of the BBB.

[0056] Preferably, the use or method is for inducing a BBB phenotype in cells in culture or in an in vitro model of the BBB.

[0057] Preferably, the use or method is for improving prediction of drug and nanoparticle delivery across the BBB based on cells in culture or in vitro models of the BBB.

[0058] Preferably, the combined use or method improves an in vitro model of the BBB.

[0059] Preferably, the combined use or method increases the tightness of the barrier formed by cells in culture or cells in an in vitro model of the BBB.

[0060] Preferably, the combined use or method promotes maturation of the barrier formed by cells in culture or in an in vitro model of the BBB.

[0061] Preferably, the combined use or method enhances brain endothelial cell identity in cells in culture or in an in vitro model of the BBB.

[0062] Preferably, the combined use or method induces a BBB phenotype in cells in culture or in an in vitro model of the BBB.

[0063] Preferably, the combined use or method improves prediction of drug and nanoparticle delivery across the BBB based on cells in culture or in an in vitro model of the BBB.

[0064] The development, e.g. improvement or acquisition, of an in vitro model of the BBB preferably comprises: Increasing the tightness of the barrier, e.g., paracellular and / or transcellular barrier, formed by cells in culture or cells in an in vitro model of the BBB; promoting maturation of the barrier formed by cells in culture or in an in vitro model of the BBB; Enhancing brain endothelial cell identity in cells in culture or in an in vitro model of the BBB; enhancing the blood-brain barrier phenotype in cells in culture or in an in vitro model of the BBB; Improving predictions of drug or nanoparticle delivery across the BBB based on cells in a cultured in vitro model of the BBB.

[0065] Preferably, the use or method of the combination of compounds comprises: Increase the tightness of the barrier, e.g., paracellular and / or transcellular barrier, formed by cells in culture or cells in an in vitro model of the BBB; promote maturation of the barrier formed by cells in culture or in an in vitro model of the BBB; Enhance brain endothelial cell identity in cells in culture or in an in vitro model of the BBB; Enhance the blood-brain barrier phenotype in cells in culture or in an in vitro model of the BBB, and / or Improve prediction of drug or nanoparticle delivery across the BBB based on cells in culture or in vitro models of the BBB.

[0066] In a further preferred embodiment of the BBB model or improved BBB model defined herein or according to the present invention, The impedance of the monolayer formed by cells in culture or in an in vitro model of the BBB is enhanced; Enhanced transendothelial electrical resistance (TEER) of monolayers formed by cells in culture or in an in vitro model of the BBB; Increased staining intensity of the tight junction protein claudin-5 in cells in culture or in an in vitro model of the BBB the permeability of the tracer compound, preferably sodium fluorescein or Evans Blue albumin, is reduced; The expression of one or more genes associated with BBB function is upregulated, down-regulation of expression of one or more genes associated with leukocyte trafficking during BBB disruption; Glycocalyx density increases, the rate of endocytosis, preferably albumin endocytosis and lipid raft / caveolin-mediated endocytosis, is reduced; increased efflux pump activity, preferably as measured by rhodamine assay; upregulated expression of one or more glycocalyx synthesis enzymes (e.g., GALNT15, ST6GALNAC1, ST6GALNAC3, ST8SIA6); down-regulated expression of one or more genes mediating endocytic vesicle formation or non-specific albumin uptake; and / or The expression of one or more genes encoding BBB efflux transporter proteins is upregulated.

[0067] In a further preferred embodiment of the BBB model or improved BBB model defined herein or according to the present invention, the effect of the combination of compounds on barrier integrity, as represented by impedance, is detectable from cells in culture after 72 hours of treatment or from cells in an in vitro model of the BBB after treatment with the combination is discontinued; The permeability of the tracer compounds sodium fluorescein or Evans blue albumin is reduced, one or more genes selected from ABCB1, ST6GALNAC3, SLC2A1 and CLDN5 are upregulated; one or more genes selected from CAV1, VCAM1, MMP2 and CCL2 are downregulated; F-actin is redistributed from stress fibers and focal adhesions to the cortical actin cytoskeleton, one or more genes selected from ABCB1, ABCG2, and ABCC4 are upregulated; Increased efflux transport of vinblastine, loperamide, salicylate, verapamil, and / or methotrexate; and / or Increased influx transport of propranolol.

[0068] Preferably, the in vitro BBB is a model of a vertebrate BBB, such as a reptile, amphibian, bird or mammal, preferably an avian or mammalian, very preferably a mammalian BBB. In particular, the BBB model is a model of the human BBB.

[0069] Preferably, the effect of the combination of compounds on barrier integrity, as represented by impedance, is detectable from cells in culture after 72 hours of treatment or from cells in an in vitro model of the BBB after treatment with the combination has ceased.

[0070] Preferably, the combination of compounds is incorporated into a kit for use in developing a model of the BBB, with at least one of the components of the combination being incorporated into a separate package.

[0071] A preferred embodiment is a preferred embodiment of any one of the aspects described herein.

[0072] a) cAMP, Ro20-1724, LiCl and A83-01; b) cAMP, Ro20-1724, Wnt-3a and A83-01; c) cAMP, Ro20-1724, LiCl and RepSox, or d) A combination consisting of cAMP, Ro20-1724, Wnt-3a and RepSox is provided.

[0073] Preferably, the combination consists of cAMP, Ro20-1724, LiCl and A83-01. [Brief explanation of the drawings]

[0074] [Figure 1]Figure 1: The combination of cAMP + Ro + LiCl + A83-01 synergistically enhances the barrier tightness of human stem cell-derived endothelial cells. a) Rationale for the study. b) Schematic of the single compounds and their combinations tested in the impedance-based screen. c) Barrier integrity of human stem cell-derived EC monolayers supplemented with PC-conditioned medium, as measured by impedance. Higher normalized cell index values ​​and larger areas under the curve indicate increased barrier integrity. Mean ± SD, ANOVA with Bonferroni's post-hoc test; ****P<0.0001 compared to control, n=6. d) Impedance kinetics of EC monolayers treated with the combination of cAMP + Ro + LiCl + A83-01, with or without the control medium change, at 72 hours. Mean ± SD, n=6. e) Schematic of a BBB model consisting of human stem cell-derived ECs acquiring brain-like properties when cocultured with PCs. f) Transendothelial electrical resistance (TEER) in the coculture model after 48 hours of treatment. Mean ± SD, ANOVA with Bonferroni's post-hoc test; ****P<0.0001 compared to the control group, n=24 from three independent experiments. g) Reproducibility of TEER measurements after 48 hours of treatment with the combination of cAMP+Ro+LiCl+A83-01 across experiments, measured by different experts using different batches of cells. Mean ± SD, ANOVA with Bonferroni's post-hoc test; n=82 from four independent experiments. h) Permeability of sodium fluorescein and Evans blue albumin in the co-culture model after 48 hours of treatment. Papp: apparent permeability coefficient. Mean ± SD, ANOVA with Bonferroni's post-hoc test; *P<0.05, **P<0.01, ****P<0.0001 compared to the control group, n=11 from two independent experiments. i) Immunostaining of the tight junction protein claudin-5 in ECs. Bar: 50 μm. Quantification: mean ± SD for middle subpanel, median ± quartile for right subpanel, analysis of variance with Bonferroni post hoc test, ****P<0.0001 compared with control group, n=27–30 from three independent experiments. [Figure 2]Figure 1: The combination of cAMP + Ro + LiCl + A83-01 promotes barrier maturation and brain endothelial cell identity as revealed by MACE-seq. a) Schematic of the experimental setup. MACE-seq was performed on samples from brain-like ECs (cocultured with PCs) treated or not with the combination of cAMP + Ro + LiCl + A83-01 for 48 hours. b) Principal component analysis (PCA) plot of control and cAMP + Ro + LiCl + A83-01-treated samples. c) Venn diagram showing the number of genes that were upregulated (log2FC ≥ 0.2 and FDR < 0.01), downregulated (log2FC ≥ -0.2 and FDR < 0.01), or unchanged (|log2FC| < 0.2 and / or FDR ≥ 0.01) upon treatment with the combination of cAMP + Ro + LiCl + A83-01. Log2FC: log2 (fold change); FDR: false discovery rate. d) Mean difference (MD) plot showing the range and distribution of all transcripts (circles) in terms of mean expression and log2FC upon treatment with the combination of cAMP + Ro + LiCl + A83-01. Key transcripts associated with different aspects of BBB function are highlighted. e) Functional enrichment analysis of up- and down-regulated gene sets using g:profiler upon treatment with the combination of cAMP + Ro + LiCl + A83-01. Gene Ontology Biological Process (GO:BP) terms are ranked based on their gene expression ratio (proportion of differentially expressed genes for a given term) and colored based on the statistical significance level (FDR). f) Representative staining of F-actin (gray) with or without treatment with the combination of cAMP + Ro + LiCl + A83-01. Nuclei were stained with Hoechst (blue). Bar: 40 μm. g) Validation of the definitive endothelial nature of human stem cell-derived brain-like ECs. Normalized counts of major epithelial and endothelial transcripts from our dataset are plotted on a three-segment y-axis covering multiple orders of magnitude. h) Scaled heatmap of 92 transcripts enriched in mouse brain ECs and well expressed in the human BBB.Genes that were upregulated, downregulated, or unchanged upon treatment with the combination of cAMP+Ro+LiCl+A83-01, as well as genes that were not expressed in this model, are shown separately in alphabetical order. Color coding represents row-wise z-scores after min-max normalization, with the darkest color corresponding to the highest expression in a given row. [Figure 3]Figure 1 shows that the combination of cAMP + Ro+LiCl+A83-01 induces a complex BBB phenotype with higher glycocalyx density, reduced endocytosis, and increased efflux pump activity. a) Schematic representation of three non-covalent aspects of BBB function induced by the combination of cAMP + Ro+LiCl+A83-01. b) Changes in gene expression of key glycocalyx synthesis enzymes at the BBB. Mean ± SD. Red indicates upregulation upon treatment with the combination of cAMP + Ro+LiCl+A83-01. c) Wheat germ agglutinin (WGA) lectin staining (gray) labels negatively charged sialic acid residues in EC glycocalyx. Nuclei were stained with Hoechst (blue). Bar: 50 μm. Quantitation (right panel): Mean ± SD, unpaired t-test, t = 11.43, df = 43, n = 22–24 from two independent experiments. d) Cell surface zeta potential measurements in ECs. Mean ± SD, unpaired t-test, t = 3.045, df = 60, n = 31 from two independent experiments. e) Changes in gene expression for mediators of endocytic vesicle formation and nonspecific albumin uptake in ECs. Mean ± SD. Blue indicates downregulation upon treatment with the combination of cAMP, Ro, LiCl, and A83-01. f) Internalization of Texas Red (TR)-labeled albumin (gray) in ECs visualized by live-cell confocal microscopy. Nuclei were stained with Hoechst (blue). Bar: 50 μm. Quantification (right panel): Mean ± SD, unpaired t-test, t = 3.786, df = 22, n = 12. g) Representative immunostaining of caveolin-1 (gray) after internalization of TR-albumin (red) in control ECs. YZ (right) and XZ (bottom) projections of the Z-stack are shown at the level of the crosshairs. Arrows point to caveolin-1+TR-albumin+ puncta, indicating TR-albumin within caveolae. Nuclei were stained with Hoechst (blue). Bar: 30 μm. h) TR-albumin internalization in ECs quantified by fluorescence spectrophotometry. MβCD: methyl-β-cyclodextrin. Incubation at 4°C inhibits the energy-dependent uptake process. Mean ± SD, two-way ANOVA with Bonferroni post-hoc test; **P<0.01, ****P<0.0001 compared to the respective 37°C group, n=6.i) Changes in gene expression of major efflux transporters at the BBB. Mean ± SD. Red indicates upregulation upon treatment with the combination of cAMP, Ro, LiCl, and A83-01. j) P-glycoprotein (P-gp, encoded by ABCB1) immunostaining in ECs (gray). Nuclei were stained with Hoechst (blue). Bar: 50 μm. Quantification (right panel): Mean ± SD, unpaired t-test, t = 4.065, df = 36, n = 19 from two independent experiments. k) Bidirectional permeability of rhodamine 123, a ligand for efflux pumps, across the BBB coculture model. Papp: apparent permeability coefficient. Mean ± SD, two-way ANOVA with Bonferroni post-hoc test, n = 4. l) Efflux ratio of rhodamine 123 calculated from the same experiment. Mean ± SD, two-way ANOVA with Bonferroni post-hoc test, n = 4. [Figure 4]Figure 1: Targeted pathways converge on Wnt / β-catenin signaling to mediate the effects of the cAMP+Ro+LiCl+A83-01 combination. a) Schematic of known interactions between effector transcription factors (TFs) of cAMP, Wnt, and TGF-β signaling. b) Changes in gene expression associated with Wnt / β-catenin signaling. Mean ± SD. Red and blue indicate upregulation and downregulation, respectively, upon treatment with the cAMP+Ro+LiCl+A83-01 combination. c) Immunostaining of β-catenin in ECs. Arrows indicate nuclear β-catenin, a hallmark of active Wnt signaling. Bar: 50 μm. Quantification: Mean ± SD for the lower left subpanel and median ± quartile for the lower right subpanel. Analysis of variance with Bonferroni post-hoc test. **P<0.01, ****P<0.0001 compared to control. n=20 from two independent experiments. d) Schematic representation of the mechanism of action of ICG-001 and XAV939, inhibitors of Wnt signaling, at different levels of the pathway. Barrier integrity is measured by impedance in EC monolayers with or without e) ICG-001 (5 μM) and f) XAV939 (1 μM). Higher normalized cell index values ​​and larger areas under the curve indicate increased barrier integrity. Mean ± SD, ANOVA with Bonferroni's post-hoc test. ****P<0.0001 compared to control, n=5-6. g) Measurement of TEER and h) permeability of sodium fluorescein and Evans blue albumin throughout the co-culture with or without ICG-001 and XAV939. Papp: apparent permeability coefficient. Mean ± SD, ANOVA with Bonferroni's post-hoc test. ****P<0.0001 compared with the respective treatment in the control group, n=6 from two independent experiments. [Figure 5]Figure 1: The combination of cAMP + Ro + LiCl + A83-01 improves in vitro prediction of drug and nanoparticle delivery across the human BBB. a) Permeability of 10 clinically used small molecule drugs in the blood-brain (AB) direction through a BBB coculture model treated with or without the combination of cAMP + Ro + LiCl + A83-01. Papp: apparent permeability coefficient; CNS: central nervous system. Red and blue colors indicate higher and lower permeability, respectively, for a given drug when treated with the combination of cAMP + Ro + LiCl + A83-01 (P<0.05). Mean ± SD for symbols, n=3-4. Lines of best fit with 95% confidence intervals and R2 values ​​from simple linear regression are shown. b) Drug efflux ratios in the coculture model. Mean ± SD, multiple unpaired t-test with Welch's correction, n=3-4. Upward-pointing red triangles and downward-pointing blue triangles indicate higher and lower efflux ratios, respectively. c) In vitro unbound brain-to-plasma partition coefficients (Kp,uu,brain) of drugs passing through the co-culture model with or without treatment with the combination of cAMP + Ro + LiCl + A83-01. Red and blue indicate higher and lower values, respectively, for a given drug when treated with the combination of cAMP + Ro + LiCl + A83-01 (P<0.05). Mean ± SD for symbols, n=3-4. Lines of best fit with 95% confidence intervals and R2 values ​​from simple linear regression are shown, respectively, when treated with the combination of cAMP + Ro + LiCl + A83-01 (P<0.05). d) Correlation heatmap of in vitro Kp,uu,brain values ​​for the entire set of drugs from our model with in vivo Kp,uu,brain data from rats and non-human primates. Numbers in boxes are Spearman's correlation coefficients. 1Data taken from this study, 2Friden et al., 2009, and 3Sato et al., 2021. e) Comparison of Kp,uu,brain measured in vivo in humans with a predictive model for verapamil. Points are means; data taken from the above papers, 4Cecchelli et al., 2014, and 5Moya et al., 2021. f) Schematic representation of non-targeted and GSH-targeted polypeptide nanoparticles (NPs) carrying rhodamine 6G cargo.g) Penetration of non-targeted NPs and GSH-targeted NPs through the co-culture BBB model measured by fluorescence spectrophotometry. Mean ± SD, two-way ANOVA with Bonferroni's post-hoc test, n=4. h) Representative live-cell confocal microscopy images showing the internalization of non-targeted NPs carrying rhodamine 6G (orange) and GSH-targeted NPs in ECs treated or not with the combination of cAMP+Ro+LiCl+A83-01. Nuclei were stained with Hoechst (blue). Bar: 50 μm. [Figure 6]The effect of the combination of cAMP + Ro + LiCl + A83-01 on barrier tightness is reproducible in two additional BBB culture models. a) Schematic representation of two widely used BBB culture models of differing complexity in which treatment with the combination of cAMP + Ro + LiCl + A83-01 was further tested. b) Real-time measurement of barrier integrity by impedance in mouse bEnd.3 cells and c) rat primary brain capillary ECs. Higher normalized cell index values ​​and larger areas under the curve indicate increased barrier integrity. Mean ± SD, analysis of variance with Bonferroni's post-hoc test; ****P<0.0001 compared to control; n=5-6 for both panels. d) Measurement of TEER and e) permeability of sodium fluorescein and Evans blue albumin across mouse bEnd.3 monolayers after 48 hours of treatment. Papp: apparent permeability coefficient. Mean ± SD, ANOVA with Bonferroni's post-hoc test. *P<0.05, ***P<0.001, ****P<0.0001 compared to the control group, n=8 from two independent experiments for both panels. Similarly, f) TEER and g) tracer permeability were measured in a coculture BBB model based on rat primary cells. Mean ± SD, ANOVA with Bonferroni's post-hoc test. ***P<0.001, ****P<0.0001 compared to the control group, n=11 from two independent experiments for both panels. h) Immunostaining of the tight junction protein claudin-5 in mouse bEnd.3 cells and i) coculture-derived rat primary brain capillary ECs. Bar: 50 μm. Quantification (right subpanel): Mean ± SD, ANOVA with Bonferroni's post-hoc test. ****P<0.0001 compared to the control group, n=20–22 from two independent experiments for both panels. [Figure 7]Figure 1 shows the effects of small molecules and recombinant proteins targeting single pathways in human stem cell-derived ECs on barrier integrity measured by impedance. Treatment concentrations and durations were optimized for a) cPT-cAMP supplemented with the cAMP-specific phosphodiesterase inhibitor Ro20-1724, b-d) activators of Wnt signaling, and e-f) inhibitors of TGF-β signaling. Concentrations selected for further experiments were 250 μM cPT-cAMP, 300 ng / mL Wnt-3a, 300 ng / mL Wnt-7a, 3 mM LiCl, 10 μM RepSox, and 3 μM A83-01. Higher normalized cell index values ​​and larger areas under the curve indicate increased barrier integrity. Mean ± SD, analysis of variance with Bonferroni post-hoc test. ****P<0.0001 compared to control; ns:P>0.05, n=6. [Figure 8] Effect of simultaneous Wnt activation and inhibition of TGF-β signaling on barrier integrity in human stem cell-derived ECs. a) Impedance kinetics of ECs treated with Wnt activators (300 ng / mL Wnt-3a, 300 ng / mL Wnt-7a, 3 mM LiCl), TGF-β receptor antagonists (10 μM RepSox, 3 μM A83-01), and their combination. Mean ± SD, n = 5. b-c) The best combination (LiCl + A83-01, Wnt-3a + RepSox) was also validated with a larger sample size. Higher normalized cell index values ​​and larger areas under the curve indicate increased barrier integrity. Mean ± SD, ANOVA with Bonferroni post-hoc test. ****P < 0.0001 compared to control, n = 10. [Figure 9]Figure 1 shows the effects of simultaneous cAMP and Wnt activation and inhibition of TGF-β signaling on barrier integrity in human stem cell-derived ECs. Impedance kinetics of ECs treated with combinations targeting all three pathways simultaneously, including a) A83-01 and b) RepSox. The same concentrations of compounds were used. Notably, some combinations (cAMP + Ro + LiCl + A83-01, cAMP + Ro + Wnt-3a + RepSox) enhanced barrier integrity more effectively than other combinations (cAMP + Ro + Wnt-3a + A83-01, cAMP + Ro + LiCl + RepSox). c) Impedance kinetics of EC monolayers treated with the two best combinations. Treatment was changed to control medium at 72 hours. Notably, the cAMP + Ro + LiCl + A83-01 combination was superior to cAMP + Ro + Wnt-3a + RepSox because its effect remained more stable upon cessation of treatment. This makes the combination of cAMP + Ro + LiCl + A83-01 a better candidate for use in permeability experiments where the presence of small molecule compounds can potentially interfere with downstream analyses. Higher normalized cell index values ​​and larger areas under the curve indicate increased barrier integrity. Mean ± SD, ANOVA with Bonferroni post-hoc test, ****P<0.0001 compared to control, n=6 for all three panels. [Figure 10]Further measurements of barrier and junctional integrity in brain-like ECs. a) Automated measurement of transendothelial electrical resistance (TEER) in brain-like ECs after treatment with the combination of cAMP + Ro + LiCl + A83-01 or its components using the cellZscope instrument. Luminal and abluminal treatments were applied only for compounds LiCl and A83-01. Compounds cAMP + Ro were always administered intraluminally. b) Manual measurement of TEER over time across the EC-PC coculture model using an EVOM volt-ohm meter and handheld STX-2 electrodes. c) Immunostaining of the tight junction protein ZO-1 in ECs from the coculture model after 48 hours of treatment. Bar: 50 μm. Quantification: Mean ± SD, analysis of variance with Bonferroni's post-hoc test; **P<0.01, ***P<0.001 compared to the control group; n=20-21 from two independent experiments. [Figure 11] Figure 1 shows gene expression changes associated with different aspects of BBB function in brain-like ECs. Gene expression changes after 48 hours of treatment with the combination of cAMP, Ro, LiCl, and A83-01 were determined using MACE-seq profiling and are shown categorically in panels a-i). Mean ± SD, n=4. Red and blue indicate up- and down-regulation, respectively, upon treatment with the combination of cAMP, Ro, LiCl, and A83-01. [Figure 12]Further experiments on Wnt / β-catenin signaling in brain-like ECs. a) Representative confocal microscopy images of VE-cadherin immunostaining (gray) show no damage to VE-cadherin junctions after 48 hours of treatment with the combination of cAMP, Ro, LiCl, and A83-01. Nuclei were stained with Hoechst (blue). Bar: 50 μm. b) Barrier integrity measured by impedance upon treatment with ICG-001, which specifically blocks β-catenin-CBP interaction in the nucleus, and c) XAV939, which blocks an upstream pathway of nuclear translocation. Cells in this experiment were not treated with the combination of cAMP, Ro, LiCl, and A83-01. The concentrations selected for further experiments were 5 μM for ICG-001 and 1 μM for XAV939. Higher normalized cell index values ​​and larger areas under the curve indicate increased barrier integrity. Mean ± SD, n = 6. [Figure 13] This figure shows that different treatment concentrations are optimal for barrier integrity in mouse bEnd.3 cells. a-f) Effect of small molecules and recombinant proteins targeting single pathways on barrier integrity, as measured by impedance. The concentrations selected for further experiments in bEnd.3 cells were 125 μM cPT-cAMP, 10 mM LiCl, 1 μM A83-01, 100 ng / mL Wnt-3a, 100 ng / mL Wnt-7a, and 1 μM RepSox. Higher normalized cell index values ​​and larger areas under the curve indicate increased barrier integrity. Mean ± SD, n = 5-6. g) Immunostaining of the tight junction protein ZO-1 in bEnd.3 cells after 48 hours of treatment. Bar: 50 μm. Quantification: Mean ± SD, analysis of variance with Bonferroni post-hoc test, n = 16 from two independent experiments. [Figure 14]Figure 1 shows that different treatment concentrations are optimal for barrier integrity in primary rat brain ECs. a-f) Effect of small molecules and recombinant proteins targeting single pathways on barrier integrity as measured by impedance. The concentrations selected for further experiments on primary rat ECs and EC-PC-AC cocultures were 250 μM cPT-cAMP, 1 mM LiCl, 1 μM A83-01, 100 ng / mL Wnt-3a, and 0.3 μM RepSox. Higher normalized cell index values ​​and larger areas under the curve indicate increased barrier integrity. Mean ± SD, n = 5-6. g) Immunostaining of the tight junction protein ZO-1 in primary rat brain ECs derived from EC-PC-AC cocultures after 48 h of treatment. Bar: 50 μm. Quantification: Mean ± SD, analysis of variance with Bonferroni's post-hoc test, n = 12 from two independent experiments. DETAILED DESCRIPTION OF THE INVENTION

[0075] A widely accepted and versatile human BBB model requires both recapitulation of the endothelial identity of the human BBB and enhancement of the barrier properties.

[0076] The term "in vitro model of the blood-brain barrier" ("blood-brain barrier model" or "BBB model") has its usual scientifically accepted meaning and refers to a construct that allows for the analysis and examination of the development and / or properties of the BBB found in animals. In vitro models of the BBB are also useful for drug evaluation. As used herein, the term "in vitro model of the blood-brain barrier" (blood-brain barrier model or BBB model) refers to a culture of cells in a layered arrangement, typically a layer on a support, preferably a porous support, where the cells communicate with other cells in the culture and exhibit one or more characteristics of the blood-brain barrier of an animal that has both blood and a brain (central nervous system), where the layer has one side representing the blood side and another side representing the brain side, and the cell culture layer controls the delivery or transport of molecules from one side to the other, particularly from the blood side to the brain side, such that certain molecules are delivered and other molecules are not delivered.

[0077] To fulfill this function, blood-brain barrier models must develop the key characteristics of BBB development: maturation of cell-cell junctions, downregulation of nonspecific vesicular transcytosis and a negatively charged glycocalyx, and efflux and influx transport systems with increased expression of specific transporters.

[0078] In a preferred embodiment, the blood-brain barrier model comprises at least endothelial cells, preferably cerebrovascular endothelial cells.

[0079] As disclosed herein, various BBB models exist, including, for example, a BBB model based on primary brain endothelial cells isolated from animal tissues, a BBB model based on stem cells such as human umbilical cord blood stem cells, and a BBB model based on induced pluripotent stem cells (hiPSCs), in which brain endothelial-like cells are differentiated from hematopoietic stem cells or induced pluripotent stem cells with appropriate differentiation factors; BBB models exist that involve the co-culture of cells such as vascular ECs with other cell types such as pericytes, astrocytes, or both; in certain variants, ECs are generated from iPSCs or CD34-positive umbilical cord blood stem cells, which are then differentiated to a brain-like phenotype by selective culture conditions and / or co-culture with brain microvascular pericytes and / or astrocytes.

[0080] As used herein, a "brain vascular endothelial cell" exhibits one or more characteristics of an endothelial cell, including (i) tight junctions between adjacent cells, (ii) the presence of solute carriers that regulate ion and small molecule transport, (iii) the expression of efflux transporters, preferably including P-glycoprotein (P-gp), breast cancer resistance protein (BCRP), and / or multidrug resistance protein (MRP), and (iv) a receptor-mediated process for the specific uptake of macromolecules, and optionally, (v) a low level of pinocytosis and (vi) the absence of fenestrae.

[0081] The abbreviation "cAMP" is herein broadly understood as any cAMP derivative that, when introduced into a cell and metabolized, provides the cell with cAMP or a cAMP analog that has the same effect as cAMP, particularly the effect of activating the cAMP / PKA signaling pathway.

[0082] The abbreviation "Ro" refers to Ro20-1724 (4-(3-butoxy-4-methoxybenzyl)imidazolidin-2-one).

[0083] Sense molecules include compounds that contain covalent bonds, however, in a broader sense, molecules are used interchangeably with compounds and can be substituted for compounds.

[0084] A "low molecular weight" compound is typically a compound whose molar mass or molecular weight (used interchangeably herein) is less than 1000 Da, preferably less than 900 Da. A low molecular weight compound may be a small molecule or a salt such as a Li salt, for example, an organic Li salt or LiCl. In a preferred embodiment, a low molecular weight compound is understood as a compound that can penetrate cells through the cell membrane without using cell surface receptors, particularly a compound with a molecular weight of less than 1000 Da, preferably less than 900 Da. In a specific embodiment, a low molecular weight compound is typically a compound with a molecular weight of less than 500 Da, thus satisfying Lipinski's "rule of five."

[0085] A compound is "cell-permeable" according to the present invention if it can be introduced (penetrate) into cells in culture, e.g., at the blood-brain barrier as defined herein. A possible method of introduction into cells is via receptors, but receptors may be expressed on the surface of cells at different levels between models or cell types, and furthermore, the "blood side" and "brain side" of a model may contain different receptors or different levels of receptors.

[0086] Therefore, in the present invention, low molecular weight compounds are preferred that can penetrate cells through the cell membrane without receptors, thus avoiding the above-mentioned variability issues and being applicable to a wide range of blood-brain barrier models.

[0087] "Comparing" two features of the BBB is understood herein to include comparing levels to determine which is higher or lower, or establishing a difference between levels or establishing a ratio of levels, or a value derived from levels, optionally completed with other mathematical procedures (calculations) required by the measurement method. In one embodiment, comparing includes subtracting two levels, e.g., subtracting two normalized or baseline-corrected levels. In one embodiment, comparing includes mathematical procedures, such as performing a transformation on both levels, e.g., calculating a logarithm or other function. In one embodiment, comparing includes generating statistics and calculating an evaluation of the statistical evaluation with a considered error and / or mean and / or statistical significance level.

[0088] As used herein, the singular forms "a," "an," and, where the context permits, "the" include plural referents unless the context otherwise dictates. Thus, where the context permits, the meaning of "a" or "an" includes the meaning "one or more."

[0089] The terms "comprises" or "comprising" or "including" herein should be construed to have a non-exhaustive meaning, allowing for the addition or inclusion of further features or method steps or components to those which include the listed features or method steps or components.

[0090] The phrases "consisting essentially of" or "comprising substantially" should be understood to mean consisting of the essential features or method steps or components recited, for example, in a claim list, but allowing for the further inclusion of other features or method steps or components that do not materially affect the essential characteristics of the use, method, composition, or other subject matter. It should be understood that "comprises" or "comprising" or "including" can be substituted herein with "consisting essentially of" or "comprising substantially," as appropriate, without adding new matter.

[0091] Selection from a list, e.g., the term "selected from" can be substituted with "selected from the group consisting of" where appropriate, and includes selecting one or more items from the list unless otherwise specified.

[0092] Treatment with the small molecule combinations described herein enhances barrier tightness in in vitro models of the BBB as measured by three independent methods (impedance, TEER (transendothelial electrical resistance), and permeability). This effect is surprisingly resilient to recovery of the small molecule combination from the culture medium, particularly in the case of cell-permeable low-molecular-weight component combinations, highly preferred cAMP, Ro, LiCl, and A83-01. It enhances barrier maturation and brain endothelial cell identity at the mRNA level, induces a complex BBB phenotype (mRNA, protein, and functional levels) with higher glycocalyx density, lower endocytosis rates, and higher activity and proper polarization of efflux pumps, acting through targeted pathways converging on Wnt / β-catenin signaling, improves in vitro predictions of drug and nanoparticle delivery across the human BBB, and reproducibly increases barrier tightness in BBB culture models.

[0093] Combinations of compounds useful in the present invention may be, for example, any combination defined in the Summary of the Invention section, particularly the first aspect of the invention, any paragraph thereof. In preferred embodiments, the compound is selected from cell-permeable low molecular weight compounds or from the specific combinations defined herein. Highly preferably, the compound is cAMP + Ro + LiCl + A83-01 (or, as a less preferred alternative, cAMP + LiCl + A83-01) or a variant thereof.

[0094] The small molecule / compound combinations described herein are useful for simultaneously improving several barrier properties in human stem cell-derived ECs or other BBB models, resulting in a more reliable BBB model for testing therapeutics that cross the BBB in vitro and for gaining mechanistic insight into BBB physiology. Furthermore, the use of the small molecule / compound combinations described herein is inexpensive, easily adaptable, efficient (synergistic effects on barrier tightness induce complex BBB phenotypes), and reproducible (in one model across experiments, across BBB models from different origins, and across BBB models of different complexity).

[0095] Over the past decade, several laboratories have established human stem cell-derived BBB models (Lippmann 2012, Cecchelli 2014, Lippmann 2014, Canfield 2017, Appelt-Menzel 2017, Praca 2019, Stebbins 2019, Qian 2019, Vatine 2019, Park 2019, Linville 2019, Missu 2019, Nishihara 2020, Moya 2021). This technology could revolutionize drug development for the brain and improve our understanding of human BBB function in health and disease. However, there is consensus in the field that existing differentiation protocols require serious improvement to faithfully recapitulate both the endothelial identity and complexity of the human BBB (Lippmann 2020, Workman 2020, Linville 2021, Lu 2021, Front Phys). Motivated by the unmet need for an easily adaptable, efficient, and reproducible approach, the inventors developed a combination of compounds or molecular cocktails that reliably induce BBB characteristics in vascular ECs. Preferred combinations and uses thereof are defined in the Summary of the Invention and the accompanying Claims. In a highly preferred embodiment, cAMP + Ro + LiCl + A83-01 has been developed.

[0096] Several methods exist for improving BBB properties in vitro. Coculture with astrocytes and / or pericytes has previously been described to enhance BBB integrity in primary brain ECs (Dehouck 1990, Nakagawa 2009). This technique has also been adapted for human stem cell-based models (Lippmann 2012, Lippmann 2014, Canfield 2017, Appelt-Menzel 2017, Praca 2019, Stebbins 2019, Park 2019, Vatine 2019, Nishihara 2020), including the one we used in this study (Cecchelli 2014). However, even with the presence of coculture signals, stem cell-derived definitive endothelial models have weak junctional tightness (Cecchelli 2014, Praca 2019, Nishihara 2020). Therefore, additional factors are needed to promote BBB maturation. Combination overexpression of a general EC-specific transcription factor and a brain EC-specific transcription factor using an adenoviral vector was recently shown to increase barrier integrity in stem cell-derived ECs by approximately 40% (Roudnicky 2020, Sci Rep). Furthermore, overexpression of three general EC transcription factors transdifferentiated cells with mixed epithelial-endothelial identity into vascular ECs that exhibit a subset of BBB characteristics (Lu 2021, PNAS). While transcriptional reprogramming is promising, it is technically challenging and relatively expensive, making it difficult to adapt for other laboratories. Similarly, selective passaging has been demonstrated to improve the precise localization of junctional proteins and reduce the permeability of fluorescein across EC monolayers (Nishihara 2020). However, this method is very time- and labor-intensive compared to other techniques. On the other hand, small molecule factors and recombinant proteins are easy to use, affordable, and, if appropriately selected, can rapidly induce BBB characteristics.

[0097] Soluble activators of cAMP (Rubin 1991, Wolburg 1994, Deli 1995), Wnt / β-catenin signaling (Paolinelli 2013, Weksler 2013, Cecchelli 2014, Laksitorini 2019, Gastfriend 2021), and retinoic acid (Lippmann 2014, Stebbins 2017) signaling, as well as inhibitors of the TGF-β (Roudnicky 2020, PNAS; Yamashita 2020) pathway, have been shown to independently enhance different aspects of BBB function in vitro. In contrast to targeting a single pathway, Praca et al. simultaneously activated multiple pathways during EC differentiation using Wnt-3α, VEGF, and retinoic acid (Praca 2019). However, the present inventors surprisingly discovered and demonstrated that cocktails containing cell-permeable small molecule compounds, such as the highly preferred combination of cAMP, Ro, LiCl, and A83-01, synergistically enhance barrier tightness in ECs via claudin-5, and mechanistically demonstrate that this effect is mediated by targeted pathways that converge on Wnt / β-catenin signaling. The combinations of the present invention induce several aspects of BBB function, which have been validated at the mRNA, protein, and functional levels. Additionally, four combinations, a subset of which possess specific advantages, as well as the highly preferred combination of cAMP, Ro, LiCl, and A83-01, have been commercialized and shown to improve in vitro predictions of drug and NP delivery across the human BBB.

[0098] Importantly, the compound combinations described herein proved applicable to a wide range of models. To this end, we focused on reproducibility using other BBB models, namely the mouse brain EC line bEnd.3 and the primary rat brain EC-pericyte-astrocyte coculture model (Nakagawa 2009). While a limitation of our study is that the interactions mediating the effects of the combinations of this invention, particularly the combination of cAMP, Ro, LiCl, and A83-01, have not yet been demonstrated in vivo, the fact that their synergistic effects are reproducible across BBB models of different species and complexity is promising. The combinations may be used in additional settings, such as fluid flow (Santa-Maria 2021), 3D microenvironments (Linville 2022), or basement membrane composition (Choi 2023), which remain to be explored.

[0099] Due to its ease of use, synergy, and reproducibility, the combinations of the present invention, particularly the highly preferred combination of cAMP, Ro, LiCl, and A83-01, have the potential to be applied to various BBB culture models across laboratories to advance drug development for the human brain. This invention is expected to be used in future studies to understand the interplay between signaling pathways that govern BBB formation, maturation, and maintenance.

[0100] Brain angiogenesis and BBB maturation are regulated by the Wnt / β-catenin pathway during development. Activation of Wnt signaling in hiPSC-derived ECs increases the BBB phenotype, but the paracellular barrier remains leaky, indicating that other signaling factors are required for complete barrier maturation. Inhibition of TGF-β signaling is a possible solution for inducing BBB properties in human stem cell-derived ECs. We hypothesized that these pathways converge and interact with cAMP signaling in ECs. Therefore, our approach was to simultaneously activate cAMP and Wnt / β-catenin and inhibit the TGF-β signaling pathway with small molecule factors to induce BBB properties in human stem cell-derived ECs (Figure 1a).

[0101] Human stem cell-derived blood-brain barrier (BBB) ​​models are powerful tools for improving our understanding of human cerebrovascular disease and facilitating drug development for the brain. However, providing endothelial cells with the appropriate molecular cues to preserve vascular identity and acquire BBB characteristics remains challenging. Here, we present a combination of cAMP, Ro, LiCl, and A83-01, an easy-to-use and affordable small-molecule cocktail that potently induces BBB characteristics in vitro. By activating cyclic AMP and Wnt / β-catenin signaling while inhibiting the transforming growth factor beta (TGF-β) pathway, the combination of cAMP, Ro, LiCl, and A83-01 synergistically enhances barrier tightness in a range of BBB models. We demonstrated that treatment with the combination of cAMP, Ro, LiCl, and A83-01 resulted in human stem cell-derived endothelial cells with a lower transcytosis rate, higher glycocalyx density, and higher efflux pump activity, accompanied by a shift in gene expression profile toward an in vivo brain endothelial signature. This invention provides mechanistic insight into how endothelial signaling is regulated during BBB maturation and leverages this to advance predictions of drug delivery to the human brain.

[0102] Their preferred combinations and uses are defined in the Summary of the Invention and in the accompanying Claims. In a highly preferred embodiment, cAMP+Ro+LiCl+A83-01 has been developed.

[0103] With regard to the improved features of the BBB model of the present invention, the following preferred embodiments are specifically taught herein: Preferably, barrier tightness is expressed as a measurement of the impedance, e.g., normalized cell index value, of the monolayer formed by cells in culture or in an in vitro model of the BBB, with higher impedance values, e.g., higher normalized cell index, indicating higher barrier tightness.

[0104] Preferably, barrier tightness is expressed as a measure of the transendothelial electrical resistance (TEER) of a monolayer formed by cells in culture or in an in vitro model of the BBB, with higher TEER values ​​for the monolayer indicating a tighter barrier.

[0105] Preferably, barrier tightness is expressed as a measure of transendothelial electrical resistance (TEER) in a co-culture of cells and pericytes in culture, with higher TEER values ​​for the monolayer indicating higher barrier tightness.

[0106] Preferably, barrier tightness is expressed as a measure of staining intensity and / or continuity of the tight junction protein claudin-5 in cells in culture or in cells in an in vitro model of the BBB, wherein increased staining intensity and / or continuity of the tight junction protein claudin-5 indicates increased barrier tightness.

[0107] Preferably, barrier tightness is expressed as a value from measurements of the permeability of a suitable tracer molecule, such as sodium fluorescein or Evans Blue albumin, with a decrease in the permeability of the suitable tracer molecule, such as sodium fluorescein or Evans Blue albumin, indicating an increase in barrier tightness.

[0108] Preferably, maturation of the barrier formed by the cells in culture or by the cells in an in vitro model of the blood-brain barrier, or enhanced brain endothelial cell identity in the cells in culture or in the in vitro model of the BBB, is measured by upregulation / increased expression of one or more genes associated with blood-brain barrier function (e.g., efflux pump proteins, glycocalyx synthesis apparatus proteins, transporter (e.g., glucose) proteins, tight junction (associated) proteins), preferably upregulation of genes selected from ABCB1, ST6GALNAC3, SLC2A1, CLDN5, indicating enhanced maturation of the barrier formed by the cells in culture or in the in vitro model of the blood-brain barrier, or enhanced brain endothelial cell identity in the cells in culture or in the in vitro model of the BBB.

[0109] Preferably, maturation of the barrier formed by the cells in culture or by the cells in an in vitro model of the blood-brain barrier, or enhanced brain endothelial cell identity in the cells in culture or in the in vitro model of the BBB, is measured by downregulation / reduction in expression of one or more genes associated with leukocyte trafficking during BBB disruption, preferably downregulation / reduction of genes selected from CAV1, VCAM1, MMP2 and CCL2, indicating enhanced maturation of the barrier formed by the cells in culture or in the in vitro model of the blood-brain barrier, or enhanced brain endothelial cell identity in the cells in culture or in the in vitro model of the BBB.

[0110] Preferably, the barrier phenotype in cells in culture or in an in vitro model of the BBB is characterized by glycocalyx density, endocytosis and / or efflux pump activity (preferably measured by a rhodamine assay), with higher glycocalyx density, reduced endocytosis rate and / or higher efflux pump activity in cells in culture or in an in vitro model of the BBB indicating an enhanced blood-brain barrier phenotype.

[0111] Preferably, the barrier phenotype in cells in culture or in an in vitro model of the BBB is characterized by the expression of one or more glycocalyx synthesis enzymes, one or more genes mediating endocytic vesicle formation or nonspecific albumin uptake, and / or one or more genes encoding BBB efflux transporter proteins, wherein increased expression of one or more glycocalyx synthesis enzymes, decreased expression of one or more genes mediating endocytic vesicle formation or nonspecific albumin uptake (e.g., PLVAP), and / or increased expression of one or more genes encoding BBB efflux transporter proteins (e.g., ABCB1, BCRP (ABCG2), MRP4, (ABCC4)) is indicative of an enhanced blood-brain barrier phenotype.

[0112] Preferably, predictions of drug or nanoparticle delivery across the blood-brain barrier based on cells in culture or cells in an in vitro model of the blood-brain barrier are measured by passive diffusion, efflux transport, or influx transport of the drug or nanoparticle, with lower efflux transport of vinblastine, loperamide, salicylate, verapamil, and / or methotrexate indicating a more reliable prediction of drug or nanoparticle delivery across the blood-brain barrier.

[0113] Furthermore, as experimental results, the following specific embodiments are disclosed herein.

[0114] Combination treatment synergistically enhances barrier tightness in human stem cell-derived endothelial cells First, we performed an impedance-based screen of small molecules and recombinant proteins for their ability to enhance barrier integrity in cultured human stem cell-derived ECs supplemented with pericyte-conditioned medium (Figure 1a). This approach allowed us to test selected activators of cAMP and Wnt, as well as inhibitors of TGF-β signaling, in real time in a 96-well plate format (Figure 1b). After testing treatment durations and concentrations (Figures 7a-f), we raised the question of whether combining Wnt activation with TGF-β inhibition would enhance barrier integrity compared to targeting a single pathway. Unexpectedly, enhancement was observed only with certain combinations, most effectively with a cell-permeable low-molecular-weight inhibitor of GSK-3-β, preferably LiCl, plus a low-molecular-weight ALK4 / 5 / 7-specific TGF-β receptor inhibitor, preferably A83-01 (Figures 8a-c).

[0115] We next investigated whether Wnt activation with LiCl or TGF-β inhibition with A83-01 could potentiate the barrier integrity-promoting effect of cAMP signaling in ECs. Targeting only the cAMP pathway with cAMP + Ro resulted in an immediate and potent increase in barrier integrity that diminished over time (Fig. 1c). Addition of LiCl or A83-01 did not potentiate the effect of cAMP + Ro (Fig. 1c). However, the combination of cAMP + Ro + LiCl + A83-01 significantly and synergistically increased barrier integrity in ECs (Fig. 1c). This effect could be reproduced by other combinations that simultaneously activated cAMP and Wnt and inhibited TGF-β signaling (Fig. 9a, b), but the combination of cAMP + Ro + LiCl + A83-01 was found to be the most potent combination (Fig. 1b, 9c). Moreover, surprisingly, the highly favorable effect of this combination of cAMP, Ro, LiCl, and A83-01 on barrier integrity was long-lasting (>72 h) and remained stable upon withdrawal of treatment from ECs after 72 h (Fig. 1d, Fig. S10a).

[0116] In initial experiments, we studied the addition of cAMP alone or together with a PDE4 inhibitor (Ro) to a human endothelial cell and brain pericyte co-culture BBB model and found that the latter effect further increased barrier integrity as measured by the normalized cell index, especially early after addition, especially within the first 24 hours (data not shown).

[0117] These findings were validated in several different BBB models, including one in which human stem cell-derived ECs acquired brain-like properties (hereafter referred to as brain-like ECs) when cocultured with pericytes (Figure 1e) (Cecchelli et al. 2014). Treatment with the combination of cAMP, Ro, LiCl, and A83-01 synergistically increased transendothelial electrical resistance (TEER) by 2.1-fold compared to the control group (Figure 1f, Figure 1b), an effect that was highly reproducible across experiments (Figure 1g). Furthermore, we measured reduced permeability of the tracers sodium fluorescein (2.2-fold) and Evans blue albumin (7.4-fold compared to the control group) through the coculture model (Figure 1h), indicating that the combination of cAMP, Ro, LiCl, and A83-01 strengthened both the paracellular and transcellular BBB. Finally, as a proposed mediator of its barrier-strengthening effect, the combination of cAMP + Ro + LiCl + A83-01 specifically (Fig. ​(Fig.1c)10c) increased both the intensity (2.3-fold) and continuity (4.5-fold) of staining for the tight junction protein claudin-5 at cell boundaries (Fig. 1i).

[0118] Combination treatment enhances barrier maturation and brain endothelial cell identity as revealed by MACE-seq To investigate gene expression changes in brain-like ECs upon treatment with the combination of cAMP + Ro + LiCl + A83-01, we performed large-scale analysis of cDNA ends (MACE-seq, Figure 2a).

[0119] Principal component analysis revealed that the control and cAMP + Ro + LiCl + A83-01-treated samples separated sharply along principal component 1 (PC1), with low heterogeneity among replicates within each group (Fig. 2b). Among the 1,054 up-regulated and 1,073 down-regulated genes after cAMP + Ro + LiCl + A83-01 treatment (Fig. 2c), we observed key genes associated with different aspects of BBB function (Fig. 2d). For example, ABCB1 (encoding the efflux pump P-glycoprotein), ST6GALNAC3 (an enzyme involved in glycocalyx synthesis), SLC2A1 (glucose transporter GLUT1), and CLDN5 (claudin 5) were up-regulated by cAMP + Ro + LiCl + A83-01 treatment. Meanwhile, CAV1 (caveolin-1) and VCAM1, MMP2, and CCL2 (vascular cell adhesion molecule-1, matrix metalloproteinase-2, and CC motif ligand-2 chemokines, respectively, all involved in pathological BBB disruption) were downregulated (Fig. 2d).

[0120] At the pathway level, we observed an overrepresentation of genes associated with EC differentiation and barrier formation, as well as maturation and quiescence (Fig. 2e), accompanied by a marked redistribution of F-actin from stress fibers to the cortical actin cytoskeleton (Fig. 2f), supporting maturation of cell-cell contacts.

[0121] Expression of key endothelial genes in brain-like ECs was several orders of magnitude higher than that of epithelial-associated genes, confirming the definitive endothelial nature of this model (Figure 2g). To investigate whether treatment with the combination of cAMP, Ro, LiCl, and A83-01 further shifted EC gene expression toward a brain endothelial signature, we examined a set of 92 transcripts that are enriched in the brain compared with peripheral ECs in mice (Vanlandewijck 2018, Sabbagh 2018) and are highly expressed in human brain capillary ECs (Yang 2022, Garcia 2022). From this list of BBB-specific, though not necessarily canonical, genes, 40 were upregulated upon treatment with the combination of cAMP, Ro, LiCl, and A83-01 (Figure 2h), indicating an incomplete but substantial shift toward in vivo brain capillary EC identity.

[0122] Combination treatment induces complex BBB phenotypes at the mRNA, protein, and functional levels MACE-seq profiling revealed that treatment with the combination of cAMP, Ro, LiCl, and A83-01 could affect not only junctional tightness but also several aspects of BBB maturation (Fig. 2d, h; Fig. 11a-i). To investigate the ability of treatment with the combination of cAMP, Ro, LiCl, and A83-01 to induce a complex BBB phenotype, we examined the protein and functional findings for three selected BBB properties in brain-like ECs (Fig. 3a).

[0123] First, the expression of key enzymes involved in the synthesis of the negatively charged EC glycocalyx was upregulated by treatment with the combination of cAMP + Ro + LiCl + A83-01 (Fig. 3b). This was reflected by the higher intensity of wheat germ agglutinin (WGA) lectin staining, which specifically recognizes negatively charged sialic acid residues in the glycocalyx, in ECs treated with the combination of cAMP + Ro + LiCl + A83-01 (Fig. 3c). Accordingly, a more negative surface charge (Δ-1.5 mV) was measured in ECs treated with the combination of cAMP + Ro + LiCl + A83-01 (Fig. 3d).

[0124] Second, genes related to endocytic vesicle formation and nonspecific albumin uptake, pathways downregulated at the BBB, were also downregulated by treatment with the combination of cAMP, Ro, LiCl, and A83-01 (Figure 3e). Expression of PLVAP (encoding plasma membrane vesicle-associated protein, a marker of peripheral ECs with high levels of endocytosis) was already low in our model and was unchanged by treatment with the combination of cAMP, Ro, LiCl, and A83-01 (Figure 3e). Treatment with the combination of cAMP, Ro, LiCl, and A83-01 resulted in an average 20–30% reduction in albumin internalization in ECs (Figure 3f), and albumin colocalized with caveolin-1 (Figure 3g). Consistent with this, albumin internalization was sensitive to methyl-β-cyclodextrin, which inhibits lipid raft / caveolin-mediated endocytosis, but to a lesser extent in cells treated with the combination of cAMP + Ro + LiCl + A83-01 (Fig. 3h).

[0125] Finally, the expression of BBB efflux transporters, such as P-glycoprotein (ABCB1), breast cancer-associated protein (BCRP, ABCG2), and multidrug resistance-associated protein 4 (MRP4, ABCC4), was upregulated by treatment with the combination of cAMP, Ro, LiCl, and A83-01 (Figure 3i). P-glycoprotein staining intensity was also 20% higher in ECs treated with the combination of cAMP, Ro, LiCl, and A83-01 (Figure 3j). As a functional test of efflux pump activity, the permeability of the efflux pump ligand, rhodamine 123, was measured throughout the coculture model in the blood-brain (AB) and brain-blood (BA) directions (Figure 3k). The efflux ratio of rhodamine 123 was 2.1 after treatment with the combination of cAMP, Ro, LiCl, and A83-01 (a 1.6-fold increase compared to the control group), and could be reduced to 0.98 by using the efflux pump inhibitor verapamil (Figure 3l). These data indicate higher glycocalyx density, reduced endocytosis, and higher activity and proper polarization of efflux pumps in brain-like ECs treated with the combination of cAMP+Ro+LiCl+A83-01.

[0126] Targeted pathways converge on Wnt / β-catenin signaling to mediate the effects of the cAMP+Ro+LiCl+A83-01 combination Next, we investigated how the synergistic effect of the combination of cAMP, Ro, LiCl, and A83-01 on barrier tightness is regulated in brain-like ECs. The STRING database proposed a network of interactions among effector transcription factors of the cAMP, Wnt, and TGF-β pathways, with β-catenin playing a central role (Fig. 4a).

[0127] Indeed, we observed a strong induction of Wnt / β-catenin signaling-related genes at the mRNA level (Fig. 4b). Furthermore, both the staining intensity and the nuclear / non-nuclear ratio of β-catenin were synergistically increased by the combination of cAMP, Ro, LiCl, and A83-01 (Fig. 4c), indicating convergence of targeted pathways for Wnt signaling. Active β-catenin was visibly present in the nucleus upon treatment with the combination of cAMP, Ro, LiCl, and A83-01 without disrupting junctional β-catenin or VE-cadherin (Fig. 4c, Fig. S12a), consistent with upregulation of RAPGEF5, a mediator of β-catenin nuclear translocation (Fig. 4b).

[0128] To investigate how the hyperactive state of Wnt / β-catenin signaling contributes to the effects of the cAMP + Ro + LiCl + A83-01 combination, we pharmacologically blocked this pathway at two different points (Fig. 4d). ICG-001 specifically inhibits β-catenin-CREB-binding protein (CBP) interaction in the nucleus, whereas XAV939 blocks a pathway upstream of nuclear translocation (Fig. 4d). These inhibitors were titrated to block the effects of high levels of Wnt signaling (Fig. 4e, f) without disrupting basal barrier integrity (Fig. 12b, c). ICG-001 moderately, and XAV939 more potently, reduced the effects of the cAMP + Ro + LiCl + A83-01 combination to nearly the level of cAMP + Ro treatment (Fig. 4e, f). Similar effects were observed when measuring TEER (Fig. 4g) and fluorescein permeability (Fig. 4h) through the coculture model. These data indicate that highly active Wnt / β-catenin signaling is required for the combined effects of cAMP+Ro+LiCl+A83-01 on barrier integrity and that β-catenin-CBP interaction is involved in establishing junctional tightness.

[0129] Combination Processing Improves Prediction of Drug and Nanoparticle Delivery Across the Human BBB To demonstrate the cAMP + Ro + LiCl + A83-01 combination, we first evaluated the permeation of 10 clinically used small molecule drugs (Tables 1 and 2 in the Examples section) through the coculture model with or without treatment with the cAMP + Ro + LiCl + A83-01 combination. Drugs were selected to encompass different permeation pathways across the BBB, including passive diffusion and efflux and influx transport. In the blood-brain (AB) direction, reduced permeation of the efflux pump ligands vinblastine, loperamide, salicylate, and verapamil was measured upon treatment with the cAMP + Ro + LiCl + A83-01 combination (Figure 5a). Conversely, the cAMP + Ro + LiCl + A83-01 combination increased the permeation of tacrine (influx) and propranolol (passive lipophilic influx) through the coculture model. As a result, the human BBB model treated with the combination of cAMP + Ro + LiCl + A83-01 was able to accurately distinguish between drugs that readily enter the central nervous system and those that do not (Figure 5a). Treatment with the combination of cAMP + Ro + LiCl + A83-01 also increased the efflux ratios of loperamide and methotrexate (with trends for vinblastine and verapamil), indicating improved directional transport across the BBB model (Figure 5b).

[0130] In vitro unbound brain-plasma partition coefficient (K p,uu,brain Similar trends were observed regarding the extent of drug permeation through the BBB model (Figure 5c), given as ). (Friden 2009, Cecchelli 2014, Moya 2021). Importantly, the combination of cAMP + Ro + LiCl + A83-01 significantly reduced the human in vitro K p,uu,brain This increased the level of correlation between the values ​​and in vivo brain penetration data from rats (Friden 2009) and non-human primates (Figure 5d) (Sato 2021). p,uu,brainIn the case of verapamil, the only drug on our list for which data were available, the BBB model treated with the combination of cAMP + Ro + LiCl + A83-01 closely approximated human in vivo values ​​(Figure 5e), second only to non-human primate in vivo data among predictive models (Sato 2021, Cecchelli 2014, Moya 2021).

[0131] To test the effect of the cAMP + Ro + LiCl + A83-01 combination on nanocarrier permeation through the co-culture model, we evaluated the permeability of polypeptide nanoparticles (NPs) with or without the BBB-specific targeting ligand glutathione (GSH; Figure 5f). Across models treated with the cAMP + Ro + LiCl + A83-01 combination, permeability relative to non-targeted NPs was 20% lower than in the control group (Figure 5g). Notably, GSH-targeted NPs exhibited increased permeability compared to non-targeted NPs in the combination-treated model but not in the control model (Figure 5g), which was also reflected by their internalization in ECs (Figure 5h). These results indicate that the cAMP + Ro + LiCl + A83-01 combination can potentially be used to improve in vitro predictions of therapeutic and NP delivery across the human BBB.

[0132] In a first aspect, the present invention provides the use of a combination of compounds in a culture of cells, the combination comprising: cAMP / PKA signaling pathway activators, a Wnt signaling pathway activator; The present invention relates to a use comprising (or consisting essentially of) a TGF-β receptor inhibitor.

[0133] Preferably, cAMP / PKA signaling pathway activator cAMP-specific PDE inhibitors, Wnt signaling pathway activators, TGF-β receptor inhibitor.

[0134] Preferably, the culture of cells is, is part of, or is present in an in vitro model of the blood-brain barrier.

[0135] In a preferred embodiment, the use of the combination of compounds is for developing an in vitro model of the blood-brain barrier (eg, for obtaining or improving an in vitro model of the blood-brain barrier).

[0136] Preferably, the cAMP / PKA signaling pathway activator is a cell-permeable, low molecular weight activator of the cAMP / PKA signaling pathway.

[0137] Preferably, the cAMP / PKA signaling pathway activator is a cell-permeable cAMP derivative (cAMP derivative) that has the effect of activating the cAMP / PKA signaling pathway, such as pCPT-cAMP, including variants such as dibutyryl-cAMP, 8-Br-cAMP, or 8-pCPT-2-O-Me-cAMP-AM.

[0138] In a preferred embodiment, the cAMP / PKA signaling pathway activator is forskolin or other direct PKA activators, such as bucladesine, virovetin, jaspermycin.

[0139] Highly preferably, the cAMP / PKA signalling pathway activator is cAMP or pCPT-cAMP.

[0140] Preferably, the cAMP-specific PDE inhibitor is a cell-permeable small molecular weight molecule.

[0141] Preferably, the cAMP-specific PDE inhibitor is a competitive cAMP-specific PDE inhibitor.

[0142] Preferably, the cAMP-specific PDE inhibitor is a cAMP-specific PDE4, PDE7 or PDE8 inhibitor, more preferably a competitive cAMP-specific PDE4, PDE7 or PDE8 inhibitor, more preferably a cAMP-specific PDE4 inhibitor, more preferably a competitive cAMP-specific PDE4 inhibitor, preferably apremilast, crisaborole or roflumilast, very preferably Ro20-1724 (4-(3-butoxy-4-methoxybenzyl)imidazolidin-2-one).

[0143] Preferably, the Wnt signalling pathway activator is a cell-permeable small molecular weight molecule.

[0144] Preferably, the Wnt signalling pathway activator is a Wnt downstream activator.

[0145] Preferably, the Wnt signaling pathway activator is a cell-permeable low molecular weight GSK-3β inhibitor, such as CHIR99021, 6-bromoindirubin-3-oxime (6-BIO); Or preferably a Li salt, a Li donor compound, a cell permeable low molecular weight Li donor compound such as Li orotate or Li carbonate, very preferably LiCl.

[0146] Preferably, the Wnt signaling pathway activator is Wnt-3a.

[0147] Preferably, the TGF-β receptor inhibitor is a cell-permeable low molecular weight molecule (small molecule).

[0148] Preferably, the TGF-β receptor inhibitor is an ALK4 or ALK5 or ALK7 (ALK4 / 5 / 7) specific TGF-β receptor inhibitor, preferably an IC 50 These are high affinity inhibitors with values ​​below the micromolar range, preferably below 500 nM, more preferably below 300 nM or below 200 nM, and very preferably below 100 nM.

[0149] Highly preferably, the TGF-β receptor inhibitor is an ALK4 / 5 / 7-specific TGF-β receptor inhibitor, such as A-83-01 (3-(6-methyl-2-pyridinyl)-N-phenyl-4-(4-quinolinyl)-1H-pyrazole-1-carbothioamide).

[0150] Preferably, the TGF-β receptor inhibitor is an ALK4 / 5 / 7-specific TGF-β receptor inhibitor, such as bactosertib, SB431542.

[0151] In a preferred embodiment, the TGF-β receptor inhibitor is an ALK5-specific TGF-β receptor inhibitor, such as SB525334, BIBF0775, very preferably RepSox; In one embodiment, the TGF-β receptor inhibitor is A-83-01 (3-(6-methyl-2-pyridinyl)-N-phenyl-4-(4-quinolinyl)-1H-pyrazole-1-carbothioamide) or RepSox.

[0152] Preferably, the combination is cAMP, a cell-permeable low molecular weight competitive cAMP-specific PDE4 inhibitor, preferably apremilast, crisaborole or roflumilast, very preferably Ro20-1724, a cell-permeable low molecular weight Li donor compound, preferably a Li salt, very preferably LiCl, and an ALK4 / 5 / 7-specific TGF-β receptor inhibitor, very preferably A83-01; cAMP, a cell-permeable low molecular weight competitive cAMP-specific PDE4 inhibitor, preferably apremilast, crisaborole or roflumilast, very preferably Ro20-1724, a cell-permeable low molecular weight Li donor compound, preferably a Li salt, very preferably LiCl, and a cell-permeable low molecular weight ALK5-specific TGF-β receptor inhibitor, very preferably RepSox, cAMP, a cell-permeable low molecular weight competitive cAMP-specific PDE4 inhibitor, preferably apremilast, crisaborole or roflumilast, very preferably Ro20-1724, Wnt-3a and a cell-permeable low molecular weight ALK4 / 5 / 7-specific TGF-β receptor inhibitor, very preferably A83-01, or The compound comprises or consists of cAMP, a cell-permeable low molecular weight competitive cAMP-specific PDE4 inhibitor, preferably apremilast, crisaborole or roflumilast, very preferably Ro20-1724, Wnt-3a and a cell-permeable low molecular weight ALK5-specific TGF-β receptor inhibitor, very preferably RepSox.

[0153] In a preferred embodiment, the cAMP is a cell-permeable cAMP analog, and in a highly preferred embodiment, pCPT-cAMP.

[0154] In one embodiment, the addition of a cell-permeable, low molecular weight competitive cAMP-specific PDE4 inhibitor, preferably apremilast, crisaborole or roflumilast, very preferably Ro20-1724, is optional.

[0155] In a highly preferred embodiment, a cell-permeable low molecular weight Li donor compound is used, preferably a Li salt, most preferably LiCl.

[0156] Preferably, the combination is The compound comprises or consists of cAMP, a cell-permeable low molecular weight competitive cAMP-specific PDE4 inhibitor, preferably apremilast, crisaborole or roflumilast, very preferably Ro20-1724, a cell-permeable low molecular weight Li donor compound, preferably a Li salt, very preferably LiCl, and an ALK4 / 5 / 7-specific TGF-β receptor inhibitor, very preferably A83-01.

[0157] Preferably, after 72 hours of treatment, the effect of the combination on barrier integrity, as indicated by impedance, is detectable after discontinuing treatment with the combination from cells in culture or cells in an in vitro model of the blood-brain barrier.

[0158] Preferably, the present invention relates to the use of combinations of small molecules where the developed BBB model remains stable upon removal of the combination from the culture medium.

[0159] Preferably, the combination is a) comprising or consisting of a Wnt signaling pathway activator selected from cAMP, Ro20-1724, LiCl, Wnt-3a, and combinations thereof, and a TGF-β receptor inhibitor selected from A83-01, RepSox, and combinations thereof.

[0160] Preferably, the combination is a) cAMP, Ro20-1724, LiCl, and a TGF-β receptor inhibitor selected from A83-01, RepSox, and combinations thereof; b) comprising or consisting of cAMP, Ro20-1724, Wnt-3a, and a TGF-β receptor inhibitor selected from A83-01, RepSox, and combinations thereof.

[0161] In one embodiment, the addition of Ro20-1724 is optional.

[0162] Particularly preferably, the combination is a) cAMP, Ro20-1724, LiCl and A83-01, and b) cAMP, Ro20-1724, Wnt-3a and A83-01; c) cAMP, Ro20-1724, LiCl and RepSox, or d) comprising or consisting of a combination selected from the group consisting of cAMP, Ro20-1724, Wnt-3a and RepSox.

[0163] Highly preferably, the combination is a) cAMP, Ro20-1724, LiCl and A83-01, and b) comprising or consisting of a combination selected from the group consisting of cAMP, Ro20-1724, Wnt-3a and A83-01; Highly preferably, the combination consists of cAMP, Ro20-1724, LiCl and A83-01.

[0164] In a highly preferred variation, the cAMP is pCPT-cAMP.

[0165] Preferably, the use is for obtaining / creating an in vitro model of the blood-brain barrier.

[0166] In a preferred embodiment, the present invention relates to the use of endothelial cells to develop in vitro models of the blood-brain barrier (eg, to obtain or improve in vitro models of the blood-brain barrier).

[0167] Preferably, the use is for improving an in vitro model of the blood-brain barrier.

[0168] Preferably, the use is for increasing the tightness of the barrier formed by cells in culture or cells in an in vitro model of the BBB.

[0169] Preferably, the use is for promoting maturation of the barrier formed by cells in culture or cells in an in vitro model of the BBB.

[0170] Preferably, the use is for enhancing brain endothelial cell identity in cells in culture or in an in vitro model of the BBB.

[0171] Preferably, the use is for inducing a blood-brain barrier phenotype in cells in culture or in an in vitro model of the BBB.

[0172] Preferably, the use is for improving prediction of drug and nanoparticle delivery across the blood-brain barrier based on cells in culture or in vitro models of the BBB.

[0173] Preferably, the combination is used to improve an in vitro model of the blood-brain barrier.

[0174] Preferably, the use of the combination increases the tightness of the barrier formed by cells in culture or in an in vitro model of the BBB.

[0175] Preferably, use of the combination promotes maturation of the barrier formed by cells in culture or in an in vitro model of the BBB.

[0176] Preferably, the use of the combination enhances brain endothelial cell identity in cells in culture or in an in vitro model of the BBB.

[0177] Preferably, use of the combination induces a blood-brain barrier phenotype in cells in culture or in an in vitro model of the BBB.

[0178] Preferably, the use of the combination improves prediction of drug and nanoparticle delivery across the blood-brain barrier based on cells in culture or in an in vitro model of the BBB.

[0179] In a preferred embodiment, the culture of cells comprises cells (optionally in an in vitro model of the blood-brain barrier) selected from the group consisting of primary brain endothelial cells (ECs), brain-like ECs, glial cells, brain pericytes, cells from brain cell lines, preferably cultures of cells comprising vascular brain ECs derived from tissue (primary) or stem cells or induced pluripotent stem cells.

[0180] The development, e.g. improvement or obtaining, of an in vitro model of the blood-brain barrier preferably comprises: Increasing the tightness of the barrier, e.g., paracellular and / or transcellular barrier, formed by cells in culture or cells in an in vitro model of the blood-brain barrier; promoting maturation of the barrier formed by cells in culture or in an in vitro model of the blood-brain barrier; Enhancing brain endothelial cell identity in cells in culture or in an in vitro model of the blood-brain barrier; enhancing the blood-brain barrier phenotype in cells in culture or in an in vitro model of the blood-brain barrier; Improving predictions of drug or nanoparticle delivery across the blood-brain barrier based on cell-cultured in vitro models of the blood-brain barrier.

[0181] The present invention also relates to a BBB model having any of these improved properties compared to a model to which the combination of the present invention is not applied. Preferably, the combination of the present invention is any one of the compound combinations defined above in the Summary of the Invention in preferred or highly preferred uses, in particular: Preferably, the use of a combination of compounds is increasing the tightness of the barrier, e.g., paracellular and / or transcellular barrier, formed by cells in culture or cells in an in vitro model of the blood-brain barrier; promote maturation of the barrier formed by cells in culture or in an in vitro model of the blood-brain barrier; Enhance brain endothelial cell identity in cells in culture or in an in vitro model of the blood-brain barrier; enhance the blood-brain barrier phenotype in cells in culture or in an in vitro model of the blood-brain barrier; and / or Improve prediction of drug or nanoparticle delivery across the blood-brain barrier based on cell in vitro models of the blood-brain barrier.

[0182] In a further preferred embodiment of the BBB model defined herein or according to the present invention, The impedance of the monolayer formed by cells in culture or in an in vitro model of the blood-brain barrier is enhanced; Enhancement of the transendothelial electrical resistance (TEER) of monolayers formed by cells in culture or in an in vitro model of the blood-brain barrier; Increased staining intensity of the tight junction protein claudin-5 in cells in culture or in an in vitro model of the blood-brain barrier The permeability of the tracer molecule, preferably sodium fluorescein or Evans Blue albumin, is reduced, the expression of one or more genes associated with blood-brain barrier function is upregulated; down-regulation of expression of one or more genes associated with leukocyte trafficking during BBB disruption; Glycocalyx density increases, the rate of endocytosis, preferably albumin endocytosis and lipid raft / caveolin-mediated endocytosis, is reduced; increased efflux pump activity, preferably as measured by rhodamine assay; upregulated expression of one or more glycocalyx synthesis enzymes (e.g., GALNT15, ST6GALNAC1, ST6GALNAC3, ST8SIA6); down-regulated expression of one or more genes mediating endocytic vesicle formation or non-specific albumin uptake; and / or The expression of one or more genes encoding BBB efflux transporter proteins is upregulated.

[0183] In a further preferred embodiment of the BBB model defined herein or according to the present invention, the effect of the combination on barrier integrity, as represented by impedance, is detectable from cells in culture after 72 hours of treatment or from cells in an in vitro model of the blood-brain barrier after treatment with the combination is discontinued; The permeability of the tracer molecule sodium fluorescein or Evans blue albumin is reduced, one or more genes selected from ABCB1, ST6GALNAC3, SLC2A1 and CLDN5 are upregulated; one or more genes selected from CAV1, VCAM1, MMP2 and CCL2 are downregulated; F-actin is redistributed from stress fibers and focal adhesions to the cortical actin cytoskeleton, one or more genes selected from ABCB1, ABCG2, and ABCC4 are upregulated; Decreased efflux transport of vinblastine, loperamide, salicylate, verapamil, and / or methotrexate; and / or Decreased transport of propranolol.

[0184] Preferably, the cells in culture or in the in vitro model of the blood-brain barrier comprise cells selected from the group consisting of primary brain endothelial cells (ECs), brain-like ECs, glial cells, brain pericytes, cells from brain cell lines, preferably cultures of cells including vascular brain ECs derived from tissue (primary) or stem cells or induced pluripotent stem cells.

[0185] Preferably, the in vitro blood-brain barrier is a model of a vertebrate blood-brain barrier, such as a reptile, amphibian, avian or mammalian, preferably an avian or mammalian, very preferably a mammalian blood-brain barrier. In particular, the BBB model is a model of the human blood-brain barrier.

[0186] Preferably, the effect of the combination on barrier integrity, as represented by impedance, is detectable from cells in culture after 72 hours of treatment or from cells in an in vitro model of the blood-brain barrier after treatment with the combination has ceased.

[0187] Preferably, the combination of compounds is incorporated into a kit for use in developing a model of the blood-brain barrier, with at least one of the components of the combination being incorporated into a separate package.

[0188] In a further aspect, the present invention also relates to a culture medium for a blood-brain barrier model as defined herein, in particular in any paragraph of the Summary of the Invention above, comprising a combination of compounds as defined in any paragraph of the Summary of the Invention, in particular as defined in the first aspect of the invention.

[0189] In a preferred embodiment, the medium is for use in developing an in vitro model of the blood-brain barrier (e.g., to obtain or improve an in vitro model of the blood-brain barrier), and the medium comprises: a cell-permeable low-molecular-weight activator of the cAMP / PKA signaling pathway, preferably pCPT-cAMP; a cell-permeable, low molecular weight, competitive cAMP-specific PDE4 inhibitor, preferably Ro-20-1724; a cell-permeable low molecular weight inhibitor of GSK-3β, preferably a cell-permeable Li-donor compound, preferably LiCl, and Preferably, it comprises a cell-permeable low molecular weight ALK4 / 5 / 7-specific TGF-β receptor inhibitor, preferably A83-01.

[0190] In a further aspect, a combination of compounds is provided, the combination comprising: cAMP / PKA signaling pathway activators, a cAMP-specific PDE inhibitor; a Wnt signaling pathway activator; and a TGF-β receptor inhibitor.

[0191] In preferred embodiments, the combination is as defined in any paragraph of the Summary of the Invention, in particular as defined in the first aspect of the invention.

[0192] In a further aspect, the present invention also provides a kit for developing a model of the blood-brain barrier, comprising: A culture of cells according to any paragraph of the Summary of the Invention, preferably comprising brain endothelial cells; a combination of compounds as defined in any paragraph of the Summary of the Invention, in particular as defined in the first aspect of the invention; and optionally an apparatus for preparing a cell culture layer, preferably comprising a porous substrate, preferably a collagen-coated porous substrate.

[0193] Preferably, the kit comprises: Brain endothelial cell cultures, and / or Brain endothelial cell and astrocyte cultures, and / or Brain endothelial cell and pericyte cultures, and / or Contains cultures of brain endothelial cells, pericytes and astrocytes.

[0194] Preferably, the combination of compounds is formulated in the form of a medium.

[0195] According to a further aspect, the present invention provides an improved blood-brain barrier model comprising: a monolayer culture of cells comprising brain endothelial cells treated with a combination of compounds as defined in the Summary paragraph, particularly the first aspect of the invention, as defined in the Summary paragraph above, compared to a blood-brain barrier model that is not treated with the compound combination. or to a model having improved characteristics as defined herein, in particular as defined in the detailed description of the invention before the experimental results, or as defined in the examples.

[0196] According to a further aspect, the present invention provides a method for developing an in vitro blood-brain barrier model (e.g., for obtaining or improving an in vitro model blood-brain barrier), comprising: preparing a layer of a culture of cells on a porous support; and culturing said cells in a medium comprising a combination of compounds as defined in the Summary paragraph of the invention, particularly the first aspect of the invention.

[0197] In a preferred embodiment, a method for developing, creating or improving an in vitro model of the blood-brain barrier (BBB) ​​is provided, comprising contacting a combination of compounds with cells in culture or in an in vitro model of the BBB; The combination of compounds is cAMP / PKA signaling pathway activators, a cAMP-specific PDE inhibitor; a Wnt signaling pathway activator; and a TGF-β receptor inhibitor.

[0198] Preferably, cells in culture or in an in vitro model of the BBB are contacted with the combination of compounds for at least 5 hours, preferably at least 10 hours, preferably at least 24 hours, preferably at least 36 hours, preferably at least 48 hours, preferably at least 72 hours.

[0199] Preferably, the cells in culture or in an in vitro model of the BBB are contacted with the combination of compounds for no longer than 72 hours.

[0200] Preferred embodiments of the first aspect are also preferred embodiments of any one of the aspects described herein.

[0201] example Materials and Methods Induction and culture of human stem cell-derived endothelial cells CD34+ hematopoietic stem cells were isolated from human umbilical cord blood and differentiated into the endothelial lineage as previously described (Cecchelli, 2014). In accordance with the World Medical Association Declaration of Helsinki, informed consent was obtained from the donor's parents, and the protocol was approved by the French Ministry of Higher Education and Research (reference number: CODECOH DC2011-1321). Endothelial cells (ECs) were derived from CD34+ stem cells in endothelial cell growth medium (EGM; Lonza, catalog number CC-3162) supplemented with 20% fetal bovine serum (FCS; Sigma-Aldrich, catalog number F7524) and 50 ng / mL vascular endothelial growth factor (VEGF165; PeproTech, catalog number 100-20). After differentiation, ECs were cultured in endothelial cell culture medium (ECM; ScienCell, Catalog No. 1001) supplemented with 5% fetal bovine serum (FBS; Sigma-Aldrich, Catalog No. F4135), 1% endothelial cell growth supplement (ECGS; ScienCell, Catalog No. 1052), and 50 μg / mL gentamicin (Sigma-Aldrich, Catalog No. G1397). For experiments, as described in detail below, ECs (passages 6–7) were expanded in coculture with bovine brain pericytes (PCs) or acquired brain-like characteristics by adding conditioned medium from bovine brain pericytes (PCs). Cells were maintained in a humidified incubator at 37 °C containing 5% CO.

[0202] Impedance-based barrier integrity measurements Real-time, label-free measurements of barrier integrity were performed in 96-well plates with integrated gold electrodes (E-Plate 96; Agilent, catalog no. 300600910) using an xCELLigence RTCA SP instrument (Agilent). Plates were coated with 0.2% gelatin (Sigma-Aldrich, catalog no. G2500), and cell-free background was measured in the medium for each well. EC was then calculated as 6 × 10 3 Cells were plated at a density of 1000 cells / well and grown to full confluence while feeding with 50% PC-conditioned medium to acquire brain-like characteristics. Confluent monolayers were incubated with 8-(4-chlorophenylthio)adenosine 3',5'-cyclic monophosphate sodium salt (cPT-cAMP; Sigma-Aldrich, Cat. No. C3912, 25-250 μM), Ro20-1724 (Sigma-Aldrich, Cat. No. 557502, 17.5 μM), recombinant human Wnt-3a protein (R&D, Cat. No. 5036-WN, 10-300 ng / mL), and recombinant human Wnt-7a protein. Monolayer impedance was measured every 5 min at 10 kHz for 48–96 h after treatment with proteins (PeproTech, catalog no. 120–31, 10–300 ng / mL), lithium chloride (LiCl; Sigma-Aldrich, catalog no. L9650, 3–30 mM), RepSox (PeproTech, catalog no. 4463325, 1–30 μM), and A83-01 (PeproTech, catalog no. 9094360, 0.1–3 μM). For combined treatments, the following concentrations were used: 250 μM cAMP, 17.5 μM Ro, 300 ng / mL Wnt-3a, 300 ng / mL Wnt-7a, 3 mM LiCl, 10 μM RepSox, and 3 μM A83-01. For all further experiments on human ECs, the cAMP + Ro + LiCl + A83-01 treatment combination consisted of 250 μM cAMP + 17.5 μM Ro + 3 mM LiCl + 3 μM A83-01. The readout of the impedance measurements is the cell index, a dimensionless parameter that correlates with the strength of cell-cell contact and cell adhesion. The cell index is Z n -Z0 is defined as Z nis the impedance at time point n, and Z is the impedance of the cell-free background in each well. Cell index values ​​were normalized to the last time point before treatment.

[0203] Construction of co-culture BBB model Primary bovine brain PCs were isolated, transfected, and maintained as previously described (Cecchelli, 2014). PCs were initially cultured on 0.2% gelatin-coated 100 mm Petri dishes in Dulbecco's modified Eagle's medium (DMEM; Gibco, catalog number 11885084) supplemented with 20% FBS (Sigma-Aldrich, catalog number F4135), 1% GlutaMAX (Gibco, catalog number 35050061), and 50 μg / mL gentamicin (Sigma-Aldrich, catalog number G1397). PC-conditioned medium from 48-h cultures was sterile filtered (0.2 μm) and used at a 1:1 ratio with EC medium to promote the acquisition of brain-like EC characteristics in experiments where coculture was not feasible. For all other experiments, a contact coculture BBB model was used. PCs were gently trypsinized and plated onto the bottom of collagen IV-coated (Sigma-Aldrich, Cat. No. C5533, 100 μg / mL) Transwell inserts (Corning, Cat. Nos. 3401 (12 mm inserts) and 3413 (6.5 mm inserts), polycarbonate membrane, 0.4 μm pore size) at 1.8 × 10 4 cells / cm 2 PCs were allowed to attach (upside down) to the bottom of the inserts for 3 hours at 37°C, and then the inserts were placed in 12- or 24-well plates containing EC medium. ECs were then placed on the other side of the inserts, which was coated with type IV collagen (100 μg / mL) and fibronectin (Sigma-Aldrich, Cat. No. F1141, 25 μg / mL), at a density of 7 × 10 in EC medium. 4 cells / cm 2Cells were seeded at a density of 0.01 mg / mL. 48 hours after co-culture establishment, the medium was replaced in both compartments. Four days after co-culture establishment, the medium was replaced in the lower compartment, and the upper compartment was treated with control EC medium or cAMP + Ro + LiCl + A83-01 or a combination of their components for 48 hours. Experiments were performed 6 days after co-culture establishment and after 48 hours of treatment with molecules or their combinations.

[0204] Transendothelial electrical resistance Transendothelial electrical resistance (TEER) was first measured across EC monolayers grown in 50% PC-conditioned medium on 6.5 mm Transwell inserts as described above using a CellZScope+ instrument (nanoAnalytics). Confluent monolayers were treated with control medium, a combination of cAMP, Ro, LiCl, and A83-01, or its components, and impedance spectra of ECs from 1 Hz to 100 kHz were monitored every 3 hours for 96 hours at 37°C. TEER was automatically calculated by the instrument and presented in arbitrary units relative to baseline. TEER was also measured across EC-PC cocultures using an EVOM volt-ohm meter (World Precision Instruments) equipped with a handheld STX2 chopstick electrode (for 12 mm inserts) or an EndOhm 6G chamber electrode (for 6.5 mm inserts). Electrodes were equilibrated in preheated EC medium for 15 minutes, and TEER was measured on a heating pad at 37°C. Ω×cm 2 The TEER at was calculated by subtracting the average resistance of two coated cell-free inserts from the raw resistance value and multiplying by the surface area of ​​the insert.

[0205] Tracer penetration After 48 hours of treatment, the upper donor compartment of the co-culture model was incubated with tracers sodium fluorescein (376 Da, Sigma-Aldrich, Cat. No. 46960, 10 μg / mL) and Evans Blue albumin complex (67 kDa, 10 mg / mL bovine serum albumin (BSA) + 167.5 μg / mL Evans Blue dye) in phenol red-free DMEM / F12 medium (Gibco, Cat. No. 21041025) supplemented with 1% FBS (Sigma-Aldrich, Cat. No. F4135). The culture was incubated at 37°C for 1 hour on a PSU-2T horizontal shaker (Biosan, 150 rpm). Samples were collected from both compartments, and the mean fluorescence of the tracers was measured using a Fluorolog3 spectrofluorometer (Horiba Jobin Yvon) at excitation / emission wavelengths of 485-515 nm for sodium fluorescein and 584 / 663 nm for Evans Blue albumin conjugate. The concentrations of the tracers were determined by standard calibration curves. The apparent permeability coefficient (Papp) was calculated using the following equation, as previously described (Veszelka, 2018):

number

[0206] Immunocytochemistry and image analysis After 48 hours of treatment, the medium was removed from the culture inserts, and ECs were fixed with ice-cold methanol-acetone (1:1 mixture) for 2 minutes (for claudin-5, ZO-1, and VE-cadherin staining) or 3% paraformaldehyde for 15 minutes at room temperature. Cells were blocked with 3% BSA in phosphate-buffered saline (PBS; 2.7 mM KCl, 1.5 mM KH2PO4, 136 mM NaCl, 6.5 mM Na2HPO4 × 2H2O, pH 7.4) for 1 hour at room temperature, followed by primary antibodies: rabbit anti-claudin-5 (Sigma-Aldrich, Cat. No. SAB4502981, AB_10753223, 1:300), rabbit anti-ZO-1 (Invitrogen, Cat. No. 61-7300, AB_2533147, 1:400), diluted in 3% BSA-PBS. The sections were incubated overnight at 4°C with rabbit anti-caveolin-1 (Santa-Cruz, catalog no. sc-894, AB_2072042, 1:100), mouse anti-P-glycoprotein (Sigma-Aldrich, catalog no. 517310, AB_564389, 1:100), rabbit anti-β-catenin (Sigma-Aldrich, catalog no. C2206, AB_476831, 1:200), and rabbit anti-VE-cadherin (Invitrogen, catalog no. PA519612, AB_10979589, 1:200). For intracellular antigens (F-actin, caveolin-1, and nuclear β-catenin), a 10-minute permeabilization step with 0.2% Triton X-100 in PBS at room temperature was performed before blocking.The cells were then incubated with the secondary antibodies Alexa Fluor 488 goat anti-rabbit IgG (Thermo Fisher, Catalog No. A-11008, AB_143165, 1:400), Alexa Fluor 555 goat anti-rabbit IgG (Thermo Fisher, Catalog No. A-21428, AB_2535849, 1:400), and Alexa Fluor 647 goat anti-mouse IgG (Thermo Fisher, Catalog No. A-21235, AB_2535804, 1:300), the F-actin probe Alexa Fluor 488 phalloidin (Invitrogen, Catalog No. A12379, 1:100), and the nuclear counterstain Hoechst 33342 (Thermo Fisher, Catalog No. H1399, 1 μg / mL) diluted in PBS for 1 h at room temperature, protected from light. Between each step, the cells were washed three times in PBS. Inserts were mounted on glass slides using Fluoromount-G (Southern Biotech, catalog no. 0100-01) and examined using a Leica TCS SP5 AOBS confocal laser scanning microscope (Leica Microsystems) equipped with HC PL APO 20x (NA = 0.7) and HCX PL APO 63x oil (NA = 1.4) objectives. Z-stacks with approximately 0.75 μm intervals were acquired. Maximum intensity projection images of each channel were analyzed by an observer blinded to the experimental groups using the "mean gray value" function in Fiji (ImageJ) software. The relative continuity of claudin-5 was quantified using MATLAB® software (The Mathworks Incorporated). To do so, images were first corrected for intensity inhomogeneity by normalized two-dimensional cross-correlation (normxcorr2) performed using a Gaussian template with a 2-pixel variance that approximates the microscope's point spread function. Image masks corresponding to junction regions were then determined by thresholding the corrected and normalized images at a threshold of 0.25. Finally, fragmentation of junction regions was quantified by calculating the ratio of the total number of distinct regions in the mask to the total mask area.Continuity was defined as the inverse of fragmentation, or simply the average size of distinct regions in the image mask. The nuclear / non-nuclear ratio of β-catenin was also determined using MATLAB® software. Regions corresponding to nuclear and cytoplasmic staining were determined by simple threshold-based segmentation performed on images of Hoechst33342-stained nuclei from the same Z-stack. The resulting segmentation masks were used to calculate the mean fluorescent pixel intensity of β-catenin staining for regions corresponding to the nucleus and cytoplasm.

[0207] Total RNA isolation Human stem cell-derived brain-like ECs (cocultured with PCs) were grown on Transwell inserts (Corning, catalog no. 3401) for 6 days. For the final 48 hours, ECs were treated with either a combination of cAMP, Ro, LiCl, and A83-01 or control medium containing the corresponding amount of solvent (diluted 100x with Milli-Q water and 1000x with DMSO) for the combination of cAMP, Ro, LiCl, and A83-01. After treatment, PCs were removed, and ECs were lysed directly on the insert using Buffer RLT Plus (Qiagen, catalog no. 1053393). Cell lysates from two inserts were pooled for one sample for further analysis. RNA was isolated using the RNeasy Plus Micro Kit (Qiagen, catalog no. 74034) with an integrated gDNA removal spin column according to the manufacturer's protocol. RNA integrity was analyzed using automated capillary electrophoresis (RNA Pico Sensitivity Assay, LabChip GX II Touch HT instrument, Perkin Elmer). Samples were stored at -80°C until further analysis.

[0208] Library preparation and 3' RNA sequencing (MACE-seq) Genome-wide gene expression profiling was performed by large-scale analysis of cDNA ends (MACE-seq). Samples containing 1 μg of purified RNA were used for library preparation. A total of eight libraries (four control and four cAMP+Ro+LiCl+A83-01 combinations) were constructed using the Rapid MACE-seq kit (GenXPro GmbH, Frankfurt, Germany) according to the manufacturer's protocol. Fragmented RNA was reverse transcribed into cDNA using barcoded oligo(dT) primers containing TrueQuant molecular barcodes, followed by template switching. Library amplification was performed using polymerase chain reaction (PCR) and purified using solid-phase reversible immobilization beads (Agencourt AMPure XP, Beckman Coulter, catalog number A63882). Sequencing was performed on an Illumina NextSeq 500 platform.

[0209] Bioinformatics analysis of MACE-seq data Approximately 58 million single 75-bp MACE-seq reads were obtained across all eight libraries. PCR duplicates were identified and removed from the raw data using TrueQuant technology. Remaining reads were further poly(A) trimmed, and low-quality reads were discarded. The clean reads were then aligned to the human reference genome (hg38, http: / / genome.ucsc.edu / cgi-bin / hgTables) using the Bowtie2 mapping tool (Langmead & Salzberg, 2012), resulting in a dataset of 21,691 genes. Gene count data were normalized to account for differences in library size and RNA composition bias by calculating median gene expression ratios using the DESeq2 R / Bioconductor package (Love, 2014). Tests for differential gene expression were also performed using DESeq2. P values ​​and log2FC (log2[fold change]) were obtained for each gene in the dataset. To account for multiple comparisons, the false discovery rate (FDR) was calculated using the Benjamini-Hochberg method. Genes with an FDR < 0.01 and |log2FC| > 0.2 were considered differentially expressed. To perform functional enrichment analysis of up- and down-regulated gene sets upon treatment with the combination of cAMP, Ro, LiCl, and A83-01, the g:GOSt tool in g:Profiler (Raudvere et al. 2019) was used to identify over-represented Gene Ontology Biological Process (GO:BP) terms and calculate gene expression ratios and statistical significance levels for each pathway. Heatmaps were generated by calculating row-wise z-scores for each gene using min-max scaling on normalized counts using the following formula:

number

[0210] Glycocalyx staining and measurement of cell surface charge After 48 hours of treatment with the cAMP + Ro + LiCl + A83-01 combination or control medium, PCs were removed from the culture inserts, and ECs were washed with PBS and fixed with 1% paraformaldehyde for 15 minutes at room temperature. Cells were then incubated with Alexa Fluor 488-conjugated wheat germ agglutinin (WGA) lectin (Invitrogen, catalog no. W11261, 5 mg / mL, 10 minutes at room temperature), which labels sialic acid and N-acetyl-D-glucosamine residues in the EC glycocalyx. Nuclei were counterstained using Hoechst 33342 (Thermo Fisher Scientific, catalog no. H1399, 1 μg / mL). Cells were washed with PBS, mounted on glass slides using Fluoromount-G (Southern Biotech, catalog no. 0100-01), and examined using a Leica TCS SP5 AOBS confocal laser scanning microscope equipped with an HCX PL APO 63x oil (NA = 1.4) objective, exactly as described above. Maximum intensity projection images of Z-stacks from each channel were analyzed using the "mean gray value" function in Fiji (ImageJ) software by an observer blinded to the experimental groups. For measurements of EC surface charge, EC-PC cocultures were established exactly as described above, but on transparent Transwell inserts (Corning, Cat. No. 3460, PET, 0.4 μm pore size), allowing for monitoring of cell detachment from the inserts. After 48 h of treatment, ECs were washed with Ringer-HEPES solution (118 mM NaCl, 4.8 mM KCl, 2.5 mM CaCl, 1.2 mM MgSO, 5.5 mM D-glucose, 20 mM HEPES, pH 7.4), gently trypsinized, and centrifuged at 700 × g for 5 min. The cell pellet was then collected using CaCl2+. 2+ and Mg 2+ The cells were resuspended in PBS supplemented with (1.2 mM CaCl2 × 2H2O, 0.5 mM MgCl2 × 6H2O) and eluted at 10 per cuvette using a Zetasizer Nano ZS instrument (Malvern) equipped with a He-Ne laser (λ = 632.8 nm) as previously described (Santa-Maria et al. 2021). 5The zeta potential of the cells was measured.

[0211] Albumin internalization ECs were cultured at 5 × 10 in 50% PC-conditioned medium on 35 mm glass-bottom Petri dishes (Greiner Bio-One, Cat. No. 627860) coated with collagen IV (Sigma-Aldrich, Cat. No. C5533, 100 μg / mL) and fibronectin (Sigma-Aldrich, Cat. No. F1141, 25 μg / mL). 4 cells / cm 2 Confluent monolayers were cultured at a seeding density of 1000 μg / mL. Confluent monolayers were treated with a combination of cAMP, Ro, LiCl, and A83-01 or control medium for 48 hours. The culture medium was then removed, and the cells were incubated with Texas Red-conjugated BSA (TR-albumin; Invitrogen, catalog no. A23017, 5 μg / mL) dissolved in control medium for 2 hours at 37°C. Hoechst 33342 (Thermo Fisher, catalog no. H1399, 2 μg / mL) was added to the cells for the final 20 minutes of incubation to counterstain the nuclei. Cultures were washed twice with Ringer-HEPES supplemented with 1% FBS, and live-cell imaging was immediately performed in the same buffer using a Leica TCS SP5 AOBS confocal laser scanning microscope equipped with an HCX PL APO 63x oil (NA = 1.4) objective, using the same parameters as above. After live-cell imaging, cells were fixed with 3% paraformaldehyde for 15 minutes at room temperature and immunostained for caveolin-1 exactly as described above. YZ and XZ projection images of Z-stacks were generated using LASX software (Leica Microsystems, version 3.7.5) to visualize TR-albumin and caveolae. To measure albumin internalization under inhibitory conditions, cells were cultured at 1.6 × 10 in 50% PC-conditioned medium in collagen IV and fibronectin-coated 24-well plates (Corning, catalog no. CLS3527, flat bottom). 4 cells / cm 2Cells were cultured at a seeding density of 1000 μg / ml. Confluent monolayers were treated with a combination of cAMP, Ro, LiCl, and A83-01 or control medium for 48 hours. To inhibit lipid raft / caveolin-mediated endocytosis, methyl-β-cyclodextrin (MβCD, Sigma-Aldrich, catalog no. C4555, 1 mM) was added to the medium of some groups for 1 hour at 37°C. Cells were then incubated with TR-albumin diluted in control EC medium containing 1% FBS for 2 hours at 37°C or 4°C. Afterwards, cells were washed three times with ice-cold PBS supplemented with 0.1% BSA and once with acid stripping buffer (50 mM glycine, 100 mM NaCl, pH 3.0) to remove cell surface-bound TR-albumin, followed by one wash with PBS. Finally, ECs were dissolved in 1% TritonX 100 in Milli-Q water, and the fluorescence signal of internalized TR-albumin was quantified using a Fluorolog3 spectrofluorometer (Horiba Jobin Yvon) at an excitation wavelength of 592 nm and an emission wavelength of 612 nm. The concentration of internalized TR-albumin was calculated using a standard calibration curve. Values ​​were normalized for the total protein content in each well using the Pierce BCA Protein Assay Kit (Thermo Fisher, catalog no. 23225) according to the manufacturer's instructions.

[0212] Efflux pump activity After 48 hours of treatment with a combination of cAMP, Ro, LiCl, and A83-01 or control medium, the permeability of the efflux pump ligand, rhodamine 123, was measured throughout the coculture model in the blood-brain (AB) and brain-blood (BA) directions. Either the luminal (blood side) or abluminal (brain side) compartment of the model was incubated with rhodamine 123 (Sigma-Aldrich, catalog no. R83702, 10 μM) diluted in phenol red-free DMEM / F12 medium (Gibco, catalog no. 21041025) supplemented with 1% FBS (Sigma-Aldrich, catalog no. F4135) for 1 hour at 37°C on a horizontal shaker set at 150 rpm. To block efflux pump activity in some groups, verapamil (Sigma-Aldrich, catalog no. V4629, 2 μM) was added to the luminal compartment 30 minutes before treatment with rhodamine 123. After 1 h, samples were collected from both compartments and the mean fluorescence of rhodamine 123 was measured at excitation / emission wavelengths of 492 / 526 nm using a Fluorolog3 spectrofluorometer (Horiba Jobin Yvon). The concentration of rhodamine 123 was determined by a standard calibration curve. The apparent transmittance coefficient (P app ) was calculated as above. P in the BA direction app The value is P in the AB direction. app By dividing by the value, the efflux ratio, which reflects the appropriate polarity of the efflux pump, was calculated.

[0213] Pharmacological inhibition of Wnt / β-catenin signaling Wnt / β-catenin signaling was blocked at two different time points using the inhibitors ICG-001 (MedChemExpress, catalog no. HY-14428) and XAV 939 (MedChemExpress, catalog no. HY-15147). To measure barrier integrity by impedance, ECs were cultured as described above. Confluent monolayers were treated with ICG-001 (5 μM–50 μM) or XAV 939 (0.5 μM–3 μM) at 37°C for 48 h, and impedance was measured every 5 min at 10 kHz using an xCELLigence RTCA SP instrument (Agilent) as described above. To examine how these inhibitors modulate the effects of treatment with the combination of cAMP, Ro, LiCl, and A83-01, ECs were co-treated with a combination of cAMP, Ro, LiCl, and A83-01 and either 5 μM ICG-001 or 1 μM XAV939 for 48 h at 37°C. To measure TEER and tracer permeability, EC-PC co-cultures were established and maintained as described above. On day 4, the luminal side (blood side) of the model was incubated with the combination of cAMP, Ro, LiCl, and A83-01 or control medium supplemented with either 5 μM ICG-001 or 1 μM XAV939 for 48 h at 37°C. TEER and tracer permeability were then measured and calculated as described above.

[0214] Permeation of small molecule drugs through BBB models The permeation of 10 selected small molecule drugs (atenolol, indomethacin, lamotrigine, loperamide, methotrexate, propranolol, salicylic acid, tacrine, verapamil, and vinblastine) was measured through the EC-PC co-culture model in the blood-to-brain (AB) and brain-to-blood (BA) directions. Drug properties and catalog numbers are listed in Table 1. Co-culture construction, maintenance, and treatment were performed as described above. After 48 hours of treatment with the cAMP+Ro+LiCl+A83-01 combination or control treatment, either the luminal (blood-side) or abluminal (brain-side) compartment of the model was incubated with 1 μM drug compound (10 mM stock prepared in DMSO) diluted in Ringer-HEPES solution (118 mM NaCl, 4.8 mM KCl, 2.5 mM CaCl, 1.2 mM MgSO, 5.5 mM D-glucose, 20 mM HEPES, pH 7.4) supplemented with 0.1% FBS (Sigma-Aldrich, catalog no. F4135) on a horizontal shaker set at 150 rpm at 37°C. Samples from both compartments were collected at 60 minutes in the BA orientation. In the AB orientation, the inserts were transferred to new wells containing fresh buffer after 30 min, and then samples were collected from the donor compartment (0–60 min) and acceptor compartment (0–30 min, 30–60 min) at the end of the assay. Samples were precipitated with methanol (Molar Chemicals, catalog number 05730-101-340) at -20 °C, and drug content was measured using a Sciex API 4000 MS-Agilent 1260 Infinity UHPLC system, while methotrexate was measured using a Sciex 500r QTOF MS-Exion LC system. Measurement parameters are listed in Table 2. Drug concentrations were determined by standard calibration curves. Multiple time points were used to measure the apparent permeability coefficient (P app ), clearance is first calculated (Nakagawa et al. 2009, Hellinger 2012, Veszelka et al., 2018),

number

number

number

number

[0215] Synthesis and penetration of nanoparticles through a BBB model Poly(L-glutamic acid) nanoparticles (NPs) were synthesized as previously described (Shen et al. 2018). Reduced L-glutathione was added to the NPs as a targeting ligand for BBB penetration at a molar ratio of L-glutamic acid:L-glutathione = 2:1 via EDC / NHS coupling. The NPs were fluorescently labeled by grafting N-(2-aminoethyl)rhodamine 6G-amide bis(trifluoroacetate) (rhodamine 6G) at a weight ratio of NP:rhodamine 6G = 1:20 via EDC / NHS coupling. The construction, maintenance, and treatment of BBB cocultures were performed as described above. After 48 hours of treatment with the combination of cAMP, Ro, LiCl, and A83-01 or control treatment, the luminal (blood-side) compartment of the model was incubated with non-targeted or GSH-targeted NPs (50 μg / mL) diluted in phenol red-free DMEM / F12 medium (Gibco, catalog no. 21041025) supplemented with 1% FBS (Sigma-Aldrich, catalog no. F4135) at 37°C on a horizontal shaker set at 150 rpm. The inserts were transferred to new wells containing fresh buffer after 1 hour, and then samples were collected from the donor compartment (0-4 hours) and acceptor compartment (0-1 hour, 1-4 hours) at the end of the assay. The mean fluorescence of rhodamine 6G was measured at excitation / emission wavelengths of 525 / -551 nm using a Fluorolog3 spectrofluorometer (Horiba Jobin Yvon). NP concentrations were determined using a standard calibration curve. Multiple time points were used to calculate the endothelial permeability coefficient (P e To calculate the β-cell count (β) (Nakagawa et al., 2009), inserts containing cells and blank inserts without cells (PS total and P.S. blank For the endothelial monolayer (PS e The permeability x surface area product values ​​of the samples were calculated using the following formula:

number

[0216] The product of the permeability of the endothelial monolayer and the surface area was then calculated by multiplying the surface area of ​​the insert available for permeability (1.12 cm 2 ) to get P e was calculated.

number

[0217] To visualize NP internalization, ECs were cultured at 5 × 10 in 50% PC-conditioned medium on 35 mm glass-bottom Petri dishes (Greiner Bio-One, Cat. No. 627860) coated with collagen IV (100 μg / mL) and fibronectin (25 μg / mL). 4 cells / cm 2 After 48 hours of treatment with the cAMP+Ro+LiCl+A83-01 combination or control medium, cells were incubated with non-targeted or GSH-targeted NPs (50 μg / mL) diluted in EC medium at 37°C for 4 hours. Hoechst 33342 (Thermo Fisher, catalog no. H1399, 2 μg / mL) was added to the cells for the final 20 minutes of incubation to counterstain the nuclei. Cultures were washed twice with Ringer-HEPES supplemented with 1% FBS, and live-cell imaging was immediately performed in the same buffer using a Leica TCS SP5 AOBS confocal laser scanning microscope equipped with an HCX PL APO 63x oil (NA = 1.4) objective, exactly as described above.

[0218] Mouse brain endothelial cell line bEnd.3 The mouse brain endothelial cell line bEnd.3 was purchased from ATCC (catalog no. CRL-2299) and cultured in DMEM / F-12 medium (Gibco, catalog no. 11320033) supplemented with 10% FBS (Sigma-Aldrich, catalog no. F4135) and 50 μg / mL gentamicin (Sigma-Aldrich, catalog no. G1397). Cells were used between passages 25 and 28. To measure barrier integrity by impedance, bEnd.3 cells were plated at 6 × 10 in rat tail collagen-coated (150 μg / mL) 96-well plates (E-plate96, Agilent, catalog no. 300600910). 3 bEnd.3 cells were seeded at a density of 6.3 × 10 cells / well and treated as described above for human ECs. To measure TEER and tracer permeability, bEnd.3 cells were plated onto Transwell inserts (Corning, Cat. Nos. 3401 and 3413, polycarbonate, 0.4 μm pore size) at a density of 6.3 × 10 cells / well. 4 cells / cm 2 bEnd.3 cells were passaged at a density of 100 μM and grown in monoculture for 4 days. bEnd.3 cells were then treated from the luminal (blood) side with a combination of cAMP, Ro, LiCl, and A83-01 optimized for this cell type (125 μM cAMP, 17.5 μM Ro, 10 mM LiCl, and 1 μM A83-01) for 48 hours at 37°C. TEER and tracer permeability were measured, and immunocytochemistry for claudin-5 and ZO-1 was performed exactly as described above for human ECs.

[0219] animal For primary cell isolation, brain tissue was collected from 3-week-old and newborn Wistar rats of both sexes (Harlan Laboratories) in accordance with the regulations of the Hungarian law of 1998 and the EU Directive 2010 / 63 / EU on animal protection and welfare. Animals were housed in a conventional animal facility at the Biological Research Centre, Szeged, Hungary, under conditions (22–24°C, 12-h light / dark cycle) with regular rodent chow and water available ad libitum.

[0220] Rat primary endothelial cell-pericyte-astrocyte coculture BBB model Isolation of primary rat brain capillary ECs, PCs, and ACs was performed as previously described (Nakagawa et al., 2009). After isolation, ECs were seeded onto 100 mm Petri dishes coated with collagen IV (100 μg / mL) and fibronectin (25 μg / mL) and supplemented with 15% plasma-derived bovine serum (PDS, First ECs were cultured in DMEM / F-12 medium (Gibco, catalog no. 11320033) supplemented with 10 mM HEPES (Sigma-Aldrich, catalog no. H4034), 1 ng / mL basic fibroblast growth factor (bFGF, Sigma-Aldrich, catalog no. F0291), 100 μg / mL heparin (Sigma-Aldrich, catalog no. 43149-100KU), 100× insulin-transferrin-selenium reagent (ITS, Gibco, catalog no. 41400045), and 50 μg / mL gentamicin (Sigma-Aldrich, catalog no. G1397). For the first 3 days of culture, EC medium also contained 3 μg / mL puromycin to eliminate P-glycoprotein-negative contaminating cell types (Perriere et al., 2005). From day 4, the PDS concentration in the EC medium was reduced to 10%. To measure barrier integrity by impedance, ECs were plated at 6 × 10 in 96-well plates (E-plate96, Agilent, catalog no. 300600910) coated with collagen IV (100 μg / mL) and fibronectin (25 μg / mL). 3 PCs were seeded at a density of 8.5 × 10 cells / well and treated as described above for human ECs. PCs were seeded on culture dishes coated with collagen IV (100 μg / mL) and cultured in DMEM (1 g / L glucose, Gibco, Catalog No. 11885084) supplemented with 10% FBS (Sigma-Aldrich, Catalog No. F4135) and 50 μg / mL gentamicin. 8.5 × 10 cells were seeded on 12-well plates (Corning, Catalog No. 3513) coated with rat tail collagen (150 μg / mL). 4 cells / cm 2ACs were seeded at a density of 1.5 × 10 cells / ml and cultured in the same medium as PCs. To construct a triple co-culture model, PCs were plated on the bottom of a Transwell insert (Corning, Cat. No. 3401, polycarbonate, 0.4 μm pore size) coated with collagen IV (100 μg / mL). 4 cells / cm 2 PCs were allowed to attach to the inserts at 37°C for 3 hours, and then the inserts were placed in a 12-well plate containing ACs, and the medium was replaced with EC medium. ECs were then plated on the other side of a culture insert coated with collagen IV (100 μg / mL) and fibronectin (25 μg / mL) at a density of 7 × 10 in EC medium. 4 cells / cm 2 The cells were seeded at a density of 100 μM. The three cell types were cultured together for 3–4 days before treatment with the cAMP+Ro+LiCl+A83-01 combination. The triple coculture model was then treated from the luminal (blood) side with the cAMP+Ro+LiCl+A83-01 combination optimized for this cell type (250 μM cAMP + 17.5 μM Ro + 1 mM LiCl + 1 μM A83-01) for 24 h at 37°C. TEER and tracer permeability were measured, and immunocytochemistry for claudin-5 and ZO-1 was performed exactly as described above for human ECs.

[0221] statistics Each individual well or insert of cultured cells that underwent the same experimental procedure was defined as a replicate. For immunocytochemistry, each individual field (4-5 fields per culture insert) was defined as a replicate. All main experiments were independently repeated, and detailed information regarding error bars, sample size, replication strategy, and statistical tests is provided in the legends of each figure. Statistical analysis was performed using GraphPad Prism software (version 5.0). Means from two or more groups were compared using unpaired t-tests or one-way analysis of variance followed by Bonferroni's post-hoc test, respectively, with normally distributed data determined by the Shapiro-Wilk test. For experiments with two independent variables, a two-way analysis of variance with Bonferroni's post-hoc test was used. The statistical significance level was set at P<0.05.

[0222] Targeted pathways converge on Wnt / β-catenin signaling to mediate the effects of combination treatment We next investigated how the synergistic effect of combined cAMP, Ro, LiCl, and A83-01 treatment on barrier tightness in brain-like ECs was regulated. The STRING database suggested a network of interactions between effector transcription factors of the cAMP, Wnt, and TGF-β pathways, with β-catenin playing a central role (Fig. 4a).

[0223] Indeed, we observed a strong induction of Wnt / β-catenin signaling-related genes at the mRNA level (Fig. 4b). Furthermore, both the staining intensity and the nuclear / non-nuclear ratio of β-catenin were synergistically increased by treatment with the combination of cAMP, Ro, LiCl, and A83-01 (Fig. 4c), indicating convergence of targeted pathways for Wnt signaling. Active β-catenin was visibly present in the nucleus upon treatment with the combination of cAMP, Ro, LiCl, and A83-01 without disrupting junctional β-catenin or VE-cadherin (Fig. 4c, Fig. S12a), consistent with upregulation of RAPGEF5, a mediator of β-catenin nuclear translocation (Fig. 4b).

[0224] To investigate how the high activity state of Wnt / β-catenin signaling contributes to the effects of cAMP + Ro + LiCl + A83-01 treatment, we pharmacologically blocked this pathway at two different points (Fig. 4d). ICG-001 specifically inhibits β-catenin-CREB binding protein (CBP) interaction in the nucleus, whereas XAV939 blocks the pathway upstream of nuclear translocation (Fig. 4d). We titrated these inhibitors to block the effects of high levels of Wnt signaling (Fig. 4e, f) without disrupting basal barrier integrity (Fig. 12b, c). ICG-001 moderately, and XAV939 more potently, reduced the effects of cAMP + Ro + LiCl + A83-01 treatment to nearly the level of cAMP + Ro treatment (Fig. 4e, f). Similar effects were observed when measuring TEER (Figure 4g) and fluorescein permeability (Figure 4h) through the coculture model. These data indicate that highly active Wnt / β-catenin signaling is required for the effects of combined cAMP, Ro, LiCl, and A83-01 treatment on barrier integrity and that β-catenin-CBP interactions are involved in establishing junctional tightness.

[0225] The effect of the combined treatment on barrier tightness is reproducible in two additional BBB culture models Finally, we investigated whether the combination of cAMP + Ro + LiCl + A83-01 could improve barrier properties in other widely used in vitro BBB models to complement the results from animal studies. To this end, we selected two models of rodent origin: the mouse brain EC line bEnd.3 in monoculture and a rat primary brain EC-pericyte-astrocyte coculture model (Figure 6a) (Nakagawa 2009). Barrier integrity, measured by impedance, was significantly increased in both models by the combination of cAMP + Ro + LiCl + A83-01, but to different degrees and with different kinetics (Figure 6b, c). Notably, the optimal treatment concentrations of the cAMP + Ro + LiCl + A83-01 combination and its components also differed between mouse bEnd.3 cells (Figures 13a–f), primary rat brain ECs (Figures 14a–f), and human brain-like ECs, highlighting the need for optimization in each model.

[0226] As validation, we measured a 2.1-fold increase in TEER (Fig. 6d) and a decrease in permeability of the tracers sodium fluorescein (1.6-fold) and Evans Blue albumin (3.4-fold) across bEnd.3 monolayers upon treatment with the combination of cAMP + Ro + LiCl + A83-01 (Fig. 6e). In a primary rat coculture model, a 2.9-fold increase in TEER (Fig. 6f) and a 2.1-fold and 1.93-fold decrease in permeability were observed for sodium fluorescein and Evans Blue albumin, respectively (Fig. 6g). Similar to human brain-like ECs, treatment with the combination of cAMP + Ro + LiCl + A83-01 synergistically and specifically increased the staining intensity of endothelial claudin-5 in mouse bEnd.3 cells (Fig. 6h, Fig. 13g) and in the primary rat coculture model (Fig. 6i, Fig. 14g). These data suggest that the conserved effect of the cAMP+Ro+LiCl+A83-01 combination across BBB models of different origin and complexity further broadens the utility of this small molecule cocktail.

[0227] [Table 1]

[0228] Table 2

[0229] Table 3

[0230] JPEG2026507796000012.jpg222153 JPEG2026507796000013.jpg228153 JPEG2026507796000014.jpg228153 JPEG2026507796000015.jpg176153

Claims

1. 1. Use of a combination of compounds in a culture of cells to develop an in vitro model of the blood-brain barrier (BBB), said combination of compounds comprising: a cAMP / PKA signaling pathway activator; a cAMP-specific PDE inhibitor; a Wnt signaling pathway activator; and a TGF-β receptor inhibitor.

2. the cAMP / PKA signaling pathway activator is a cell-permeable low-molecular-weight activator of the cAMP / PKA signaling pathway; the cAMP-specific PDE inhibitor is a cell-permeable, low-molecular-weight competitive cAMP-specific PDE4 inhibitor; the Wnt signaling pathway activator is a cell-permeable, low-molecular-weight Wnt signaling pathway activator, preferably a cell-permeable, low-molecular-weight inhibitor of GSK-3β; and / or The use according to claim 1, wherein the TGF-β receptor inhibitor is a cell-permeable low molecular weight ALK5 TGF-β receptor inhibitor.

3. The use according to claim 1 or 2, wherein the Wnt signalling pathway activator is a cell-permeable low molecular weight inhibitor of GSK-3β, preferably a cell-permeable Li-donor compound, preferably LiCl.

4. The TGF-β receptor inhibitor is a cell-permeable low molecular weight ALK4 / 5 / 7-specific TGF-β receptor inhibitor, preferably IC ５０ The use according to any one of claims 1 to 3, wherein the value is in the range of less than 500 nM, preferably in the range of less than 200 nM, preferably wherein said TGF-β receptor inhibitor is A83-01.

5. The combination of compounds is pCPT-cAMP, Ro20-1724, LiCl, and A83-01; pCPT-cAMP, Ro20-1724, Wnt-3a, and A83-01; pCPT-cAMP, Ro20-1724, LiCl and RepSox, or The use according to any one of claims 1 to 4, comprising or consisting of pCPT-cAMP, Ro20-1724, Wnt-3a and RepSox.

6. The use according to claim 5, wherein the combination of compounds consists of pCPT-cAMP, Ro20-1724, LiCl and A83-01.

7. 7. The use of claim 6, wherein the effect of the combination of compounds on barrier integrity, as represented by impedance, is detectable from cells in the culture after 72 hours of treatment and after treatment with the combination is discontinued.

8. Use of the combination of compounds enhances the tightness and / or maturity of the barrier formed by cells in the culture. and / or enhancing brain endothelial cell identity and / or BBB phenotype in the cells in said culture; and / or 9. The use of any one of claims 1 to 8, which improves prediction of drug or nanoparticle delivery across the BBB based on cells in culture.

9. 7. A culture medium for a cell-based BBB model comprising a combination of compounds according to any one of claims 1 to 6.

10. The medium according to claim 9, comprising pCPT-cAMP, Ro20-1724, LiCl and A83-01.

11. 7. An improved BBB model comprising a monolayer culture of cells treated with the combination of compounds according to any one of claims 1 to 6.

12. A kit for developing a BBB model, comprising: a culture of cells, preferably comprising brain endothelial cells; A combination of compounds according to any one of claims 1 to 6. and optionally an apparatus for preparing a cell culture layer, preferably comprising a porous substrate, preferably a collagen-coated porous substrate.

13. The use according to any one of claims 1 to 8, the culture medium according to claim 9 or 10, the improved BBB model according to claim 11, or the kit according to claim 12, wherein the cells are selected from the group consisting of brain endothelial cells, primary brain endothelial cells (ECs), brain-like ECs, glial cells, brain pericytes, cells derived from brain cell lines, vascular brain ECs, primary vascular brain ECs, vascular brain ECs derived from induced pluripotent stem cells, and induced pluripotent stem cells.

14. The cells brain endothelial cells, or Brain endothelial cells and pericytes, or brain endothelial cells and astrocytes, or 14. The use, medium, improved BBB model or kit according to claim 13, comprising brain endothelial cells, pericytes and astrocytes.

15. In the cell, The staining intensity of the tight junction protein claudin-5 increased, The permeability of the tracer molecule, preferably sodium fluorescein or Evans Blue albumin, is reduced, expression of one or more genes associated with the function of the BBB is upregulated; down-regulating the expression of one or more genes associated with leukocyte trafficking during BBB disruption; Glycocalyx density increases, the rate of endocytosis, preferably albumin endocytosis and lipid raft / caveolin-mediated endocytosis, is reduced; increased efflux pump activity, preferably as measured by rhodamine assay; upregulated expression of one or more glycocalyx synthesis enzymes; down-regulated expression of one or more genes mediating endocytic vesicle formation or non-specific albumin uptake; and / or expression of one or more genes encoding BBB efflux transporter proteins is upregulated; and / or The impedance of the monolayer formed by the cells in culture is enhanced, The use of any one of claims 1 to 8 and 12 to 14, or the improved BBB model of any one of claims 11 to 14, wherein the transendothelial electrical resistance (TEER) of the monolayer formed by the cells in culture is enhanced.

16. the effect of the combination of compounds on barrier integrity, as represented by impedance, is detectable in cells in the culture after 72 hours of treatment and after treatment with the combination is discontinued; The permeability of the tracer molecule sodium fluorescein or Evans blue albumin is reduced, One or more genes selected from ABCB1, ST6GALNAC3, SLC2A1 and CLDN5 are upregulated; one or more genes selected from CAV1, VCAM1, MMP2 and CCL2 are downregulated; F-actin is redistributed from stress fibers and focal adhesions to the cortical actin cytoskeleton, one or more genes selected from ABCB1, ABCG2, and ABCC4 are upregulated; Increased efflux transport of vinblastine, loperamide, salicylate, verapamil, and / or methotrexate; and / or The use according to any one of claims 1 to 8 and 12 to 15, or the improved BBB model according to any one of claims 11 to 15, wherein the influx transport of propranolol is increased.

17. 16. The use or improved BBB model of claim 15, wherein the effect of the compound combination on barrier integrity, as represented by impedance, is detectable from cells in the culture after 72 hours of treatment and after treatment with the compound combination has ceased.

Images