Thymic constructs and uses thereof

Abstract

The present invention provides a method of producing a thymic construct suitable for implantation into a subject, the method comprising the steps of: (i) providing a decellularised scaffold; (ii) seeding the decellularised scaffold with thymic epithelial cells having mesenchymal properties; and (iii) culturing the seeded scaffold to produce the construct. The thymic epithelial cells are preferably CD49f + , also VIM + , TE-7 + , and / or CD90 + . The present invention further provides pharmaceutical compositions, uses and therapies using thymic epithelial cells having mesenchymal properties.

Description

TECHNICAL FIELD

[0001] The present invention relates to thymic constructs, methods of manufacturing thymic constructs and uses thereof. BACKGROUND

[0002] The thymus is a major lymphoid organ in which hematopoietic progenitor cells are instructed to become functional T cells. The thymus undergoes postnatal involution but remains present throughout life.

[0003] The epithelial component of the thymic stroma is derived from the endoderm, while the mesenchymal stroma is derived from the neural crest, with a mesenchymal stroma portion derived from the afferent blood vessels. To reflect their fundamental role in immune regulation, many studies have delved into the origin and phenotypic complexity of thymic epithelial cells (hereinafter "TECs") and the molecules that mediate their activity, including AIRE and MHC class II (MHCII) required to select T cells with different non-self invasion specificities.

[0004] A central, long-established TEC distinguishing feature is their assignment as cortical (cTEC) or medullary (mTEC), reflecting the anatomical region to which differentiated TECs contribute. However, despite rather detailed studies on defined TECs and other stromal cell types, only incomplete specific molecular and cellular understanding is available on how TECs and other cells collectively contribute to the complex three-dimensional (3D) organization of the human thymus.

[0005] As the thymus controls the development of immune competence and tolerance, its functional dissection and subsequent reconstitution with desired cell populations can provide a powerful tool applicable to many diseases, including primary or acquired immune deficiencies. Indeed, cultured thymic lobes are currently used in clinical demand for transplantation into athymic patients, and many adverse complications exist.

[0006] Furthermore, thymic therapeutic strategies can be equally applicable to organ transplantation, where tolerance is a key obstacle to long-term graft acceptance.

[0007] Attempts to reconstitute a fully functional thymus have so far been met with limited success, possibly because of the inherent complexity of the thymus. Moreover, the properties of postnatal epithelial stem / progenitor cells and mesenchymal cells, including the ability to extensively expand in vitro, remain largely uncertain.

[0008] It would therefore be advantageous to provide a method of constructing a functional thymus using a thymic tissue scaffold.

[0009] It would also be advantageous to provide a method of producing a thymic tissue construct from a decellularized thymus.

[0010] Furthermore, it would be desirable to be able to produce thymic tissue devices with phenotypically replicating thymocytes, so that the resulting construct is as closely matched as possible in phenotype to the native in vivo thymus.

[0011] Furthermore, it would be advantageous to provide a single thymocyte type that is able to repopulate a decellularised scaffold, particularly as this can be used without the need to add HSCs.

[0012] Thus, it is an object of embodiments of the application to overcome or alleviate at least one problem of the prior art, whether or not this is disclosed herein. SUMMARY

[0013] According to a first aspect of the application, there is provided a method of producing a thymic construct suitable for implantation into a subject, the method comprising the steps of:

[0014] (i) providing a decellularised scaffold;

[0015] (ii) seeding thymic epithelial cells to the decellularised scaffold; and

[0016] (iii) culturing the seeded scaffold to produce the construct.

[0017] The decellularised scaffold can be a decellularised tissue scaffold or a synthetic scaffold. Such scaffolds and methods of their production are well known in the art. For example, WO0214480 describes a number of scaffold categories: (1) non-degradable synthetic polymers; (2) degradable synthetic polymers; (3) non-human collagen gels, which are non-porous; (4) non-human collagen meshes, which are processed to the desired porosity; and (5) decellularised tissues.

[0018] The decellularised scaffold typically does not include cells or cellular components. However, it will be appreciated that, for example, where a scaffold from a biological source is used (e.g. a decellularised scaffold), some cells can remain on the scaffold (e.g. following decellularisation), as described below.

[0019] In one embodiment herein, the scaffold is an artificial scaffold, which can be a synthetic or a natural polymer scaffold.

[0020] Other synthetic scaffolds can be proteinaceous in nature, e.g. consisting primarily of purified proteins such as collagen. Non-synthetic scaffolds can also be proteinaceous, or consist primarily of the collagen extracellular matrix (ECM) in tissue.

[0021] The scaffold can be a 3D printed scaffold, which can comprise any of the aforementioned materials.

[0022] Preferably, the scaffold is a decellularised (bio)matrix.

[0023] In some embodiments, the scaffold comprises a decellularized thymus scaffold.

[0024] In preferred embodiments, the decellularized thymus scaffold is a decellularized whole thymus scaffold, preferably a decellularized whole thymus; and in preferred embodiments, the resulting construct is a reconstituted thymus. In other embodiments, the decellularized thymus scaffold is a decellularized portion of a thymus, such as a lobule or gland of a thymus.

[0025] The inventors have surprisingly found that specific thymic epithelial cells (TECs) can be used to generate a functional thymus when seeded and cultured on a decellularized scaffold. Surprisingly, TECs exhibit epithelial-mesenchymal mixed cell functions, are able to expand long-term, and are able to reconstitute the anatomical phenotype of a native thymus when combined with a decellularized extracellular matrix (ECM) of a thymus or other suitable scaffold. This anatomical human thymus reconstitution is functional as judged by its ability to support the development of mature T cells upon transplantation into human immunodeficient mice.

[0026] The resulting construct can be considered an artificial organ, in particular an artificial thymus.

[0027] TECs suitable for seeding onto a decellularized scaffold exhibit mesenchymal characteristics, properties and / or functions. By "mesenchymal characteristics" we mean that the TECs comprise one or more expression markers present in mesenchymal cells, in particular thymic mesenchymal cells. By "mesenchymal properties" and "mesenchymal functions" we mean that the TECs exhibit mesenchymal behavior and epithelial behavior and are able to differentiate into different thymic cell types.

[0028] Surprisingly, it has been found that such a "hybrid" population of TECs with both epithelial and mesenchymal functions can be isolated and used to create a thymus construct without the need for further cell seeding with other cell types.

[0029] In some embodiments, the population of TECs exhibiting mesenchymal properties is isolated, purified and expanded prior to step (ii). Isolation, purification and expansion can be performed by any suitable method routine in the art. Thus, the TECs are preferably clonally derived TECs.

[0030] It has been found that the use of an isolated and expanded population of cells provides for a more efficient repopulation of the decellularized scaffold compared to known methods in which thymic cells are simply digested and applied to the decellularized scaffold or in which the TECs are not isolated prior to culturing the thymic cell population.

[0031] In some embodiments, the TECs comprise medullary TECs (mTECs), cortical TECs (cTECs), or preferably a combination of mTECs and cTECs, in particular at least one of mTECs and cTECs, and preferably both mTECs and cTECs, exhibit mesenchymal expression markers.

[0032] In other embodiments, the TECs can be derived from precursors or stem cells, for example induced pluripotent stem cells that have been induced to differentiate into TECs, in particular mTECs and cTECs.

[0033] In some embodiments, the TECs express CD49f (CD49f+). In some embodiments, the TECs are CD49f+, and at least one selected from CD90+, VIM+, and TE-7+.

[0034] In preferred embodiments, the TECs are substantially all CD49f+. In particularly preferred embodiments, the CD49f+ TECs are also VIM+ and optionally TE-7+.

[0035] In preferred embodiments, the TECs are substantially all CD49f+. In particularly preferred embodiments, the CD49f+ TECs are also VIM+ and optionally TE-7+.

[0036] In some embodiments, the TECs express CD90 (CD90+) and CD49f (CD49f+).

[0037] In preferred embodiments, the TECs express CD49f (CD49f+) CD90 (CD90+) and VIM (VIM+).

[0038] In preferred embodiments, the TECs are CD49f+ CD90+ VIM+ and TE-7+.

[0039] In some embodiments, the TECs are substantially all CD90+. In some embodiments, the CD90+ TECs also express VIM and optionally TE-7.

[0040] Surprisingly, the present inventors found that a high proportion (>60%) of mTECs and substantially all cTECs that exhibit mesenchymal characteristics express the mesenchymal markers CD49f and CD90 (Thyl), and thus such selected TECs exhibit a hybrid epithelial-mesenchymal phenotype.

[0041] This application also surprisingly discovers that isolated mTECs and / or cTECs expressing CD49f (CD49f+) and VIM (VIM+) and / or CD90 (CD90) possess the ability to be passaged and significantly expanded in vitro. Therefore, isolated (preferably clonal) CD49f+, CD90+, CD90+CD49f+, CD49f+VIM+, and / or CD90+VIM+CD49f+ epithelial cells can be individually aggregated for subsequent seeding and refilling of decellularized thymus scaffolds, proliferation, and differentiation to form an artificial thymus. Particularly surprising is that CD49f+, CD90+, VIM+, and TE-7+ cells can be seeded onto decellularized scaffolds and differentiate to generate viable thymus constructs suitable for implantation in subjects, as many other TEC populations do not exhibit such properties or functions.

[0042] In a preferred embodiment, step (ii) includes seeding a decellularized thymic scaffold with thymic epithelial cells and thymic stromal cells (TICs). The thymic stromal cells (TICs) may include one or more cell types selected from mesenchymal cells, fibroblasts, and pericytes; and may include any combination of two or all of the above cell types.

[0043] The decellularized thymus scaffold can be co-implanted with TEC and TIC.

[0044] The TEC and TIC ratio can be between 3:1 and 10:1, preferably between 4:1 and 8:1, and most preferably around 5:1.

[0045] Although seeding a decellularized scaffold with TEC alone yields constructs in which the TEC is reorganized along the 3D extracellular matrix (ECM) structure from the subcapsular region to the innermost region of the scaffold and supported by the scaffold's 3D ECM, the reconstructed scaffold containing thymic epithelial cells and thymic stromal cells creates a more functional microenvironment that supports the development of human in vitro lymphoid progenitor cells and in vivo hematopoietic stem cell (HSC) T cells. Thymic constructs generated by seeding TEC, or TEC and TIC, do not require any separate mesenchymal cell seeding or any other stem cell or precursor cell treatment.

[0046] In other embodiments, the TEC can be seeded with non-thymic mesenchymal cells.

[0047] The TEC and TIC can be human TEC and TIC. Human TEC and TIC can be obtained from a living human donor before step (ii) or before becoming a cadaver.

[0048] In embodiments, wherein the decellularised scaffold is a decellularised thymus or a portion of a decellularised thymus, the thymus or portion of a thymus can be a human thymus or portion of a human thymus obtained from a living human donor or a cadaver.

[0049] Thus, by overcoming the obstacles to constructing a functional thymus with only expanded stromal cells, the present application is believed to provide a practical prospect for treating immune diseases including congenital athymia, for which current treatments are limited.

[0050] The production method of the present application, and the thymus construct produced thereby, is typically performed in vitro. However, it is contemplated that further cell proliferation and / or differentiation and production of the construct can occur following implantation in vivo. Thus, it is preferred that the production of the construct is performed in vitro until a construct is produced that is sufficiently populated with TECs that are sufficiently differentiated to allow successful implantation into a subject.

[0051] Subsequently, for example following implantation, further cell proliferation can occur within and / or on the scaffold. Thus, it is contemplated that the scaffold need not be fully populated with seeded cells for implantation into a subject. For example, the scaffold can have regions where no seeded cells are present, for example the scaffold can have seeded cells over at least 70%, 80%, 90%, 95% or at least 99% of the area of its surface.

[0052] The cells used in the method are typically autologous, i.e. derived from or intended recipient of the tissue or organ construct produced by the method of the present application. However, the cells used in the method can also be allogeneic, i.e. derived from a subject other than the non-recipient subject from which the tissue or organ construct is produced. Furthermore, xenogeneic cells can be used, i.e. cells derived from a tissue / organ construct recipient of a different species. The cells can also be produced from pluripotent stem cells or induced pluripotent stem cells.

[0053] Preferably, seeding of the TECs (and TICs, when present) is performed by injection into the scaffold or perfusion through the vasculature of the scaffold. Injection can be performed in multiple stages. If a whole thymus scaffold is used, seeding can be performed by injection into a portion of the thymus scaffold, e.g. one or more lobes, or by injection into multiple sites of the thymus scaffold. The cells can be delivered individually, simultaneously or sequentially. Sequential delivery can comprise delivering populations of cells within at least 1, 2, 5, 10, 20, 30, 40, 50 or 60 minutes of each other, or within 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 14, 36 or 48 hours of each other.

[0054] TECs or TECs and TICs can be injected into and / or onto each lobe of the decellularized thymic scaffold at 1-10M, 1-8M, 2-6M, or 2-3M cells per 1 to 1000 μΐ of media per lobe.

[0055] The seeded scaffold is typically maintained in vitro to allow the cells to migrate and initially differentiate on the reticulated ECM of the scaffold for at least 1, 2, 3, or at least 4 days, up to 21, 14, 12, 10, 9, 8, or 7 days, preferably about 4-7 days. Donor cells, such as stem cells and / or hematopoietic cells, can then be delivered into or onto the construct, preferably by injection.

[0056] The decellularized ECM scaffold or seeded ECM scaffold can be additionally treated to enhance cell adhesion to the scaffold and / or cell growth on the scaffold. Such treatment can include the application of proteins such as growth factors or extracellular matrix proteins, for example, including one or more of collagen, elastin, fibronectin, laminin, or proteoglycans.

[0057] In some embodiments, step (ii) can include seeding additional thymic cells other than TECs and optionally TIC cells. Other suitable thymic cells include thymic endothelial cells, fibroblasts, and hematopoietic stromal cells, such as dendritic cells.

[0058] A "decellularized" scaffold typically does not contain cells or cellular components. However, it is contemplated that, for example, in the case of a scaffold from a biological source (e.g., a decellularized scaffold), some cells can remain on the scaffold (e.g., after decellularization), as described below.

[0059] In some embodiments, the decellularized scaffold comprises a decellularized thymic tissue, preferably a decellularized whole thymus or a decellularized whole thymic lobe.

[0060] In one embodiment, a human neonatal donor thymic scaffold can be used. In another embodiment, the scaffold can be derived from a human cadaver. In yet another embodiment, the scaffold can be derived from a living patient.

[0061] The scaffold can be xenogeneic, i.e., it is derived or derived from a donor of a different species than the recipient, e.g., a human recipient.

[0062] In this regard, suitable decellularized scaffolds include decellularized animal-derived scaffolds, e.g., porcine-derived, rat-derived, or rabbit-derived scaffolds. For example, in preferred embodiments, the scaffold can be a decellularized rat, porcine, sheep, or rabbit thymus, more preferably a porcine thymus.

[0063] Any suitable known decellularisation method can be employed to provide the scaffold. Typically, the decellularisation method employs various chemical, biochemical and / or physical means to disrupt, degrade and / or destroy cellular components and / or modify the matrix in which the cells are embedded to facilitate removal of the cells and cellular components, typically leaving an ECM scaffold. WO0214480 describes methods of decellularising natural tissue. The present application includes the use of a decellularised scaffold produced by any decellularisation technique which substantially removes all cells whilst leaving the ECM substantially intact. By leaving the ECM "substantially intact" is meant that at least 70, 80, 90, 95, 99 or substantially 100% of the matrix, e.g. ECM, is retained.

[0064] In one embodiment, the decellularisation of the thymic scaffold is carried out by perfusing the tissue with at least one decellularisation medium.

[0065] Suitable decellularisation media include detergents, such as sodium dodecyl sulphate (SDS), sodium deoxycholate (SOC), detergents comprising a hydrophilic polyoxyethylene and a hydrophobic hydrocarbon moiety, such as Triton X-100 (RTM); enzymes, such as proteolytic enzymes, e.g. trypsin; and nucleases, such as deoxyribonucleases, e.g. deoxyribonuclease I, and ribonucleases, e.g. RNase, and combinations thereof. Tributylphosphine phosphate (TBnP) can also be included in one or more of the decellularisation media. TBnP is a solvent which disrupts protein-protein interactions.

[0066] The method preferably comprises perfusing the tissue with more than one decellularisation medium. Suitably, the method comprises perfusing the tissue with at least one detergent and at least one nuclease. In some embodiments, the method comprises steps of perfusing the tissue with a detergent and perfusing the tissue with a nuclease, respectively, and the detergent perfusion step can be carried out prior to the nuclease perfusion step. Following perfusion and decellularisation, the decellularised thymic scaffold can be irradiated with gamma rays; and can be stored at 1-4°C ready for use.

[0067] Each perfusion step can be carried out at a temperature between 15°C and 45°C, or between 20°C and 40°C. When the perfusion step comprises an enzyme perfusion step, this can be carried out at around 37°C, which is particularly advantageous when the perfusion step comprises a nuclease material. When the perfusion step is a detergent perfusion step, this can be carried out at a temperature between 15°C and 40°C, for example at around ambient temperature or room temperature.

[0068] In other embodiments, the decellularisation can comprise subjecting the thymic tissue or thymus to a permeabilisation shock treatment.

[0069] According to a second aspect of the application, there is provided an isolated thymic epithelial cell for use in therapy.

[0070] The thymic epithelial cells can be used for seeding onto a thymic scaffold for constructing an artificial thymus or for repairing damaged thymic tissue.

[0071] In some embodiments, there is provided a combination of thymic epithelial cells and thymic mesenchymal cells for use in thymus regeneration.

[0072] According to a third aspect of the application, there is provided an isolated thymic epithelial cell for use in the treatment of a thymic, immune or autoimmune disease or a thymic congenital defect.

[0073] According to a fourth aspect of the application, there is provided a pharmaceutical composition comprising an isolated thymic epithelial cell and a pharmaceutically acceptable carrier.

[0074] The thymic epithelial cells of the second, third and fourth aspects of the application can be as described above for the first aspect of the application, and in particular can be TECs exhibiting mesenchymal as well as epithelial, features, properties and / or functions; and can be CD49f+, or CD49f+and at least one selected from CD90+, VIM+and TE-7+. In some embodiments, the TECs express CD90 (CD90+) and CD49f (CD49f+). In preferred embodiments, the CD49f+ TECs are also VIM+and optionally TE-7+. In some embodiments, the TECs express CD90 (CD90+) and CD49f (CD49f+). In preferred embodiments, the TECs express CD49f (CD49f+) CD90 (CD90+) and VIM (VIM+). In preferred embodiments, the TECs are CD49f+ CD90+ VIM+and TE-7+. In some embodiments, the TECs are substantially CD90+. In some embodiments, the CD90+ TECs also express VIM and optionally TE-7.

[0075] In some embodiments, the TECs comprise medullary TECs (mTECs), cortical TECs (cTECs), or preferably a combination of mTECs and cTECs, in particular a combination of mTECs and cTECs at least one of which, preferably both, exhibit mesenchymal expression markers.

[0076] In other embodiments, the TECs can be derived from stem cells, such as induced pluripotent stem cells, which have been induced to differentiate into TECs, in particular mTECs and cTECs.

[0077] The TECs are preferably a mixture of isolated mTECs and isolated cTECs, preferably a mixture of isolated, purified and expanded mTECs and cTECs.

[0078] The pharmaceutical composition can further comprise isolated thymic interstitial cells (TICs).

[0079] The pharmaceutical composition can comprise a TEC:TIC ratio between 3:1 to 10:1, preferably between 4:1 to 8:1, most preferably around 5:1.

[0080] In some embodiments, the TECs are CD49f+or CD49f+and at least one selected from CD90+, VIM+, and TE-7+. In some embodiments, the TECs express CD90 (CD90+) and CD49f (CD49f+). In preferred embodiments, the CD49f+TECs are also VIM+and optionally TE-7+. In some embodiments, the TECs express CD90 (CD90+) and CD49f (CD49f+). In preferred embodiments, the TECs express CD49f (CD49f+), CD90 (CD90+), and VIM (VIM+). In preferred embodiments, the TECs are CD49f+, CD90+, VIM+, and TE-7+. In some embodiments, the TECs are essentially CD90+. In some embodiments, the CD90+TECs also express VIM and optionally TE-7. In some embodiments, all TECs in the pharmaceutical composition express at least CD49f+, and essentially all TECs in the pharmaceutical composition can comprise CD49f+and / or CD90+thymic epithelial cells and a pharmaceutically acceptable carrier.

[0081] The pharmaceutical composition of the fourth aspect of the application can be used for the regeneration of the thymus or for the treatment of a thymic, immune or autoimmune disease.

[0082] According to a fifth aspect of the application, there is provided a thymic construct obtained or obtainable by the method of the first aspect of the application. The thymic construct can be an artificial thymus. The thymic construct can be used for therapy; and can be used for implantation into a subject in need of a thymic construct implantation, for example a subject with a damaged or absent thymus.

[0083] According to a sixth aspect of the application, there is provided a method of treating a thymic, immune deficiency or autoimmune disease or deficiency in a subject in need of such treatment, comprising implanting into said subject a thymic construct as defined in the fifth aspect of the application above. The disease can be athymia.

[0084] Preferably, the subject is a mammal, and more preferably a human.

[0085] Thus, in all aspects of the invention, long-term expanded epithelial cells display a hybrid epithelial-mesenchymal phenotype hitherto unprecedented. These cells can repopulate entire organ scaffolds, phenotypically mimicking the unique 3D epithelial network of the thymus. It has been found that this reconstituted scaffold can establish a functional microenvironment that supports T cell development of human lymphoid progenitors in vitro and hematopoietic stem cells (HSCs) in vivo.

[0086] The isolated cells, constructs and artificial organs can be used in therapy and can be used in surgical methods comprising implanting the construct, artificial organ or cells into a patient.

[0087] According to a seventh aspect of the invention, there is provided a method of decellularising a thymus, the method comprising:

[0088] a. providing a thymus or a portion of a thymus or a thymic lobe;

[0089] b. occluding all arteries except one artery entering the thymus, portion of a thymus or thymic lobe; and

[0090] c. perfusing decellularisation media into the thymus, portion of a thymus or thymic lobe through the remaining artery.

[0091] In some embodiments, the carotid artery is left open. In preferred embodiments, the left carotid artery is left open.

[0092] In preferred embodiments, the arteries of the thymus, portion of a thymus or thymic lobe are occluded in the following order: right common carotid artery, right subclavian artery, right internal mammary artery, right costocervical trunk, right aortic arch, left aortic arch, left costocervical trunk, and left internal mammary artery, with the left common carotid artery left open for subsequent cannulation and organ perfusion. BRIEF DESCRIPTION OF DRAWINGS

[0093] Embodiments of the present invention will now be described, by way of example only, with reference to the accompanying drawings in which:

[0094] Figure 1A 、 1B and 1C illustrate the results of isolation of thymic epithelial cells (TECs) and mesenchymal cells (TMCs). (A) FACS plots showing expression of CD90 in EpCAM high -mTECs, EpCAM low -cTECs and EpCAM - TICs, respectively. Gating was determined according to negative unstained control values for each channel. Same thymus as in 1 b. (B) Representative FACS analysis of CD49f and CD90 showing CD49f + (type 1) and CD49f -(C) Rhodamine-stained sorted cTEC and mTEC subtypes (4000 / well) were plated in indicated wells for 12 days, and colony forming efficiency (CFE). mTEC Type1 are the most clonal population (2-4%), cTEC Type1 (1-2%) are clonal population of the cortex (n=4).

[0095] Figures 2a to 2i show the results of whole organ thymus perfusion and decellularization. (a) Gross appearance of cannulated rat thymus before (upper panel) and after (lower panel) decellularization. A 24G cannula was inserted into the carotid artery and the organ was perfused with detergent and enzyme solutions. The asterisk (*) indicates the extra-thymic tissue that allowed the connection between the thymic tissue and the cannula through a large blood vessel (n=120). Scale bar, 2 mm. (b) Micro-CT image of a cannulated rat thymus showing the extra-thymic tissue, the large blood vessels and the 24G cannula into the artery. Iodine contrast shows a clear boundary between the cortex (C, bright) and medulla (M) regions; the blood vessels (*) are represented by very bright areas between and inside the thin-walled tissue (n=3). Scale bar, 1.5 mm. (c) Micro-CT 3D image of a cannulated whole rat thymus with iodine contrast showing the vascular tree (segmented in red, n=2). Scale bar, 1.2 mm. (d) Masson's trichrome staining of a fresh rat thymic lobe, with red keratin, blue collagen and pink cytoplasm staining. C is the cortex and M is the medulla (n=3 thymuses). Scale bar, 250 pm. (e) Hematoxylin and eosin (H&E) staining of a fresh rat thymus. C is the cortex and M is the medulla (n=3 thymuses). Scale bar, 500 pm. (f) Micro-CT image of a cannulated rat thymus showing a 3D image of the complete 2 lobes. Scale bar, 1.2 mm. (g) Micro-CT image of a cannulated rat thymus injected with GelMa and thresholded to show perfusion of the two thymic lobes. Scale bar, 1.2 mm. (h) Masson's trichrome staining of a decellularized rat thymus scaffold paraffin section showing collagen fibers (blue), absence of keratin, muscle fibers and cytoplasm (n=3 scaffolds). Scale bar, 250 pm. (i) H&E of a decellularized thymus scaffold showing preservation of the complete thymic lobular ECM and large and small blood vessel walls (n=4 scaffolds). Scale bar, 500 pm.

[0096] Figures 3A to 3HResults showing whole organ thymic scaffold functional repopulation. (A) Gross microscopy of thymic scaffolds before injection (left panel), shortly after injection of stromal cells (middle panel) and after 4 days of culture (right panel). The density from the empty thymic lobes was gradually increased and remodeled by the contraction of the scaffold and the increase in tissue volume (n = 60 re-populated scaffolds). Scale bar, 4 mm. (B) Immunofluorescence labeling of thymic epithelial cells (TECs) grown within the decellularized scaffold shows the presence of CK5 / 14 + , TPR63 + and CD49f + TECs. Nuclei are stained with DAPI (n = 4 re-populated scaffolds). Scale bar, 30 μm. (C) H&E staining shows the histology of scaffolds re-populated with TECs only and cultured for 5 days (n = 4 re-populated scaffolds). Scale bar, 50 μm. (D) Hematoxylin and eosin (H&E) staining of scaffolds re-populated with clonally expanded TECs and thymic interstitial cells (TICs) and cultured for 5 days prior to fixation and histological analysis. Stromal cells reorganize along the scaffold, a pattern similar to that observed in (E) early (9 weeks post-conception, wpc) human fetal thymus (n = 4 re-populated scaffolds and n = 2 human fetal thymuses). Scale bar, 100 μm. (F) Immunofluorescence labeling of TECs seeded together with TICs and grown within the decellularized scaffold shows that CK5 / 14 + cells are localized in subcapsular regions, whereas CK8 + cells are generally localized in internal regions; TPR63 + TECs are predominantly CD49f+. Nuclei are stained with DAPI (n = 4 re-populated scaffolds). Scale bar, 100 μm. (G) H&E staining of scaffolds re-populated with TECs, TICs and hematopoietic progenitor cells after 7 days of culture (n = 4 re-populated scaffolds). Scale bar, 100 μm. (H) Representative FACS analysis of CD45-positive populations isolated from re-populated scaffolds seeded with triple negative (TN, CD3 - CD4 - CD8 - ) progenitor cells and co-cultured for 8 days (n = 6 re-populated scaffolds in three independent experiments). FSC-A, SSC-A plots show the presence of cells as well as debris derived from the scaffold ECM during the dissociation process to release hematopoietic cells (top left panel). Live cells represent approximately 90% of total cells (top middle panel). TNs formed within the scaffold give rise to double positive (DP) and single positive (SP) CD4 and CD8 expressing cells (top right panel, 5000 cells). Live cells are positive for CD1 a and negative for CD33 (bottom right panel). CD4 and CD8 are positive for CD3 and express TCRaβ (middle and bottom left panels).

[0097] Figure 4 Results showing that the refilled thymic scaffold matured in vivo and promoted functional T cell development in the form of H&E staining of histological sections of thymic scaffolds harvested at different time points post-transplantation in mice (8, 11, 18 and 22 wpt). (*) indicates Hassall’s bodies (HB); n = 18 scaffolds in 3 independent experiments. Scale bar, 100 pm. DETAILED DESCRIPTION

[0098] Method

[0099] Animals

[0100] All animal procedures were in accordance with ethical approval and Home Office Project Licences (PPL). (NSG) and NOD.CgFoxn1 em1Dvs .Prkdc scid .Il2Rγc tm1Wjl (NSG nude mice, stock number: 026263) were obtained from Jackson Laboratories. Mice were maintained in a 12-hour light-dark cycle with an ambient temperature of 19 / 22 °C and humidity of 45 / 65%. Sprague-Dawley (SD) rats were housed in a breeding facility.

[0101] Human tissue

[0102] Postnatal thymus was donated by patients (age 3 days to 11 years) undergoing cardiothoracic surgery at Great Ormond Street Hospital, London, UK.

[0103] Preparation and culture of clinical thymus sections

[0104] Following removal of the capsule, 8-15 grams of postnatal thymus tissue was processed using a Stadie Riggs microtome (Thomas Scientific) to obtain 1 mm thick sections. Sections were mounted individually on nitrocellulose filters (Millipore), subsequently placed on Spongostan surgical sponge (Ferrosan Medical Devices) and immersed in culture medium (F12 (Gibco), 10% heat-inactivated fetal bovine serum (Gibco) and 1% penicillin-streptomycin (Sigma)) in a 10 cm Petri dish for 21 days; culture medium was changed daily.

[0105] CD34 + HSC purification and sorting

[0106] Mononuclear cells (MNC) from two to five CB collections were pooled and purified by Ficoll-Paque density centrifugation (GE Healthcare Life Sciences, Buckinghamshire, UK) followed by ammonium chloride red blood cell lysis. Density-isolated CB MNC were subjected to CD34-positive magnetic sorting by EasySep Human CD34 Positive Selection Kit (Stemcell Technologies) according to the manufacturer’s instructions; or they were sorted by fluorescence-activated cell sorting (FACS). Fetal liver (FL) CD34 + positive cells were isolated from human fetal tissue at 12 to 20 weeks post-conception. Tissue digestion was performed using an enzyme solution (0.1 U / mL collagenase A (Roche), 0.8 U / mL dispase II (Roche), and 100 pg / mL DNase I (Roche) in RPMI, 2% FBS, and 1% penicillin / streptomycin solution) at 37°C. Cells were pelleted and processed for ammonium chloride red blood cell lysis. FL MNC were subjected to CD34-positive magnetic sorting.

[0107] Thymocyte isolation and sorting

[0108] Human thymocytes were obtained by mechanical tissue dissociation of postnatal thymus. Triple negative (TN, CD3 - CD4 - CD8 - ) cells were enriched by magnetic sorting using biotinylated anti-CD3, anti-CD4, anti-CD8a, and anti-CD235ab antibodies (BioLegend) and Magnisort SAV negative beads (Invitrogen). The negatively selected fraction was processed for FACS purity sorting (CD11c - CD19 - CD56 - CD3 - CD4 - CD8 - cells; FACSAria III machine, BD Bioscience; FACSDiva 8.0.1 software).

[0109] Isolation and culture of thymic stromal cells

[0110] Human thymuses were obtained from 33 patients aged between 3 days and 11 years old who underwent open-heart cardiac surgery. All thymuses, regardless of donor age, contained TECs that produced clonal colonies that could extensively expand after weekly passaging. Thymus tissue fragments were dissociated into single cells by enzymatic treatment (0.4 mg / mL collagenase D (Roche), 0.6 mg / mL dispase II (Roche), and 40 pg / mL deoxyribonuclease I (Roche)) for about 30-45 min at 37°C. Cells were pelleted and resuspended for cell counting. Freshly dissociated thymus cells were plated on a layer of lethally irradiated mouse fibroblasts (3T3-J2). Alternatively, thymus single-cell suspensions were depleted of cells expressing CD45 and CD235ab (erythrocytes) by immunomagnetic separation. Prior to cell culture on 3T3-J2 feeder cells, cortical (EpCAM low CD205 + ) and medullary (EpCAM high CD205 - ) thymic epithelial cells (TECs) were stained and sorted. Epithelial cells were expanded in cFAD medium composed of 3:1 DMEM-1X (Gibco) and F-12 NutMix (Gibco) with the addition of 10% fetal bovine serum (Sigma), 1% penicillin and streptomycin (100X, Sigma), corticosterone (0.4 pg / ml, Calbiochem), cholera enterotoxin (10 -10 M, Sigma), triiodothyronine (T3) (2 x 10 -9 M Sigma), and insulin (5 pg / ml, Sigma). Human epithelial growth factor (hEGF, 10 ng / ml, PeproTech) was added after three days of culture and then every other day at feeding time. TECs were plated at a density of 2000-4000 cells / cm 2 and incubated at 37°C and 6% CO2, reaching confluence between 5 and 7 days of culture. TEC colony- forming efficiency (CFE) assays were performed once every other passage. 500 cells were obtained by serial dilution and plated on 60 mm tissue culture dishes on lethally irradiated 3T3-J2 feeder cells. After 12 days of culture, growing colonies were fixed in 4% paraformaldehyde (PFA) and stained with rhodamine-B (1%, SIGMA-ALDRICH) for 15 min. Colonies were scored under a dissecting microscope. Thymic interstitial cells (TICs) were sorted by depletion of EpCAM -Thymic populations of stromal cells or obtained by explants of human thymus. Thymic tissue fragments were placed on 60 mm culture dishes previously coated with Matrigel (BD Biosciences) at 1 : 100 dilution in Megacell medium (Sigma). Fragments were adhered for 30 minutes and then gently overlaid with Megacell medium supplemented with 2.5% FBS HI (Life Technologies), 1% Penicillin / Streptomycin (Sigma), 1% L-glutamine (Life Technologies), 1% non-essential amino acids (Life Technologies), 100 mM β-mercaptoethanol (Life Technologies) and basic fibroblast growth factor (Sigma). After 7 days of culture, cells grown from the explants were harvested and cultured at 37°C in a gas mixture of 6% CO2 and 5% O2, and passaged every 3-4 days when TIC reached confluence. TM (Corning) at 1 : 100 dilution in Megacell medium (Sigma). Fragments were adhered for 30 minutes and then gently overlaid with Megacell medium supplemented with 2.5% FBS HI (Life Technologies), 1% Penicillin / Streptomycin (Sigma), 1% L-glutamine (Life Technologies), 1% non-essential amino acids (Life Technologies), 100 mM β-mercaptoethanol (Life Technologies) and basic fibroblast growth factor (Sigma). After 7 days of culture, cells grown from the explants were harvested and cultured at 37°C in a gas mixture of 6% CO2 and 5% O2, and passaged every 3-4 days when TIC reached confluence.

[0111] Lymphocyte culture and in vitro stimulation experiments

[0112] Human CD45 + / CD3 +Sorted cells were cultured in 96, 48 and 24 well plates (Falcon) in RPMI 1640 (Gibco), Glutamax (Life Technologies) and 10% fetal bovine serum (FBS, Life Technologies), 1% penicillin-streptomycin (Life Technology), and added with human recombinant IL-2 (50 U / mL, R&D system) and IL-7 (5 ng / mL, Invitrogen) for 28 days. To expand, anti-CD3 / CD28 beads (Thermofisher) were added at a 1:1 ratio. Cytokine production was assessed by phorbol myristate acetate (PMA) and ionomycin (Io) stimulation assay: human CD45+ / CD3+ sorted T cells and freshly isolated human thymocytes were incubated at 37°C for 6h in RPMI added with 20% human serum (heat inactivated) (SIGMA-ALDRICH), protein transport inhibitor mixture (500X, eBioscience), PMA (40 ng / mL, SIGMA-ALDRICH) and Io (4 pg / mL, SIGMA-ALDRICH). T cells were then washed with HBSS solution and stained for CD3, CD4 and CD8 (BioLegend), stained with APC-Cy7 fixable live dye (Invitrogen), then fixed and permeabilized with intracellular staining buffer kit (BioLegend) and intracellularly stained with anti-IFNy, anti-TNFa and anti-IL-2 antibodies (BioLegent). Expanded and unstimulated T cells were used to set FACS gates.

[0113] Endothelial cells (HuVEC VeraVec) culture

[0114] Human endothelial cells (Vascular Biologics, Cat. No. HVERA101) were cultured on a 0.1% gelatin (Merck) layer of culture medium 199 (Gibco) added with 20% fetal bovine serum, FBS (Gibco), 1% antibiotic / antimycotic (ThermoFisher), 10 mM HEPES (Gibco), 100 pg / ml heparin (SIGMA-ALDRICH) and 50 pg / ml endothelial supplement ECGS (Millipore).

[0115] Flow cytometry analysis

[0116] Single-cell suspensions were stained on ice for 30 minutes with a specific antibody mixture (Supplementary Table 1) in Hanks balanced salt solution (HBSS, Life Technologies) with 2% FBS (Life Technologies) added. DAPI (SIGMA-ALDRICH) or Zombie live / dead cell dye (Invitrogen) was used to distinguish between live and dead cells. A Fortessa X-20 instrument (BD FACSDiva 8.0.1 software) and FlowJo were used. TM The software (BDBioscience) was used for FACS phenotypic analysis.

[0117] RNA isolation and RT-qPCR

[0118] Collect cultured and freshly isolated cells in ReliaPrep. TM Gene expression analysis was performed using either the BL buffer from the Promega kit or the Trizol TRI reagent (SIGMA-ALDRICH). The precipitated and dried RNA was resuspended in nuclease-free water (Qiagen). RNA concentration was determined using a Nanodrop 1000 (ThermoScientific), and RNA integrity (RIN) was assessed using a BioAnalyzer 2100 (Agilent). If necessary, follow the manufacturer's instructions. RNA amplification system V2 (Nugen) amplifies RNA. Alternatively, according to the manufacturer's specifications, GoScript can be used. TM A reverse transcriptase kit (Promega) was used to convert RNA into DNA. The cDNA concentration was adjusted to 10 ng / μl. Quantitative (q)PCR was performed using a QuantStudio 3 real-time PCR system (Applied Biosystems) in MicroAmp rapid optical 96-well plates with PCR premix (PrecisionPLUS-R-Primerdesign Ltd) and low-ROX and Taqman qPCR probes (Integrated DNA Technology, Supplementary Table 2).

[0119] Single-cell RNA sequencing – 10X genomics

[0120] Cell numbers were confirmed using an Eve automated cell counter (NanoEnTek). Where possible, appropriate volumes containing 10,000 cells were topped up to 46.6 μΐ with nuclease-free water. For lower cell concentrations, 46.6 μΐ of cell suspension was loaded without further dilution. Reverse transcription and library construction was performed according to the Chromium Single Cell 3’ Reagent v3 protocol (10x Genomics) according to the manufacturer’s recommendations. Total complementary DNA synthesis was performed using 12 amplification cycles, with final cDNA yields of approximately 3 ng / μΐ to 15 ng / μΐ. 10x Genomics sequencing libraries were constructed as described and sequenced on an Illumina HiSeq 4000 with read lengths of 28-8-98.

[0121] Bioinformatic data analysis of single cell data

[0122] Raw sequencing data was processed using the CellRanger pipeline (10x Genomics). Count tables were loaded into R and further processed using the Seurat 3R package 59 We removed all observed cells with less than 200 unique genes, or with more than 20% (samples mTEC1, mTEC2) or 30% (samples cTEC1, cTEC2) unique molecular identifiers originating from mitochondrial genes. Principal component analysis was then performed on the apparently variable genes, and based on manual inspection of the principal component variance plot (“PC scree plot”), the top 20 principal components were selected for clustering and UMAP. Clustering was performed using the default method in the Seurat package, with a resolution parameter set to 0.5. Area under the curve (AUC) summary intensity plots were prepared using the R-package AUCell.

[0123] Preparation and analysis of bulk RNA sequencing libraries

[0124] Total RNA was extracted from cultured cells using RNeasy Micro Kit (Qiagen) and RNA total concentration was measured using a Nanodrop spectrophotometer. rRNA was removed from total RNA with Ribo-Zero, reverse transcribed using Superscript II Reverse Transcriptase (Invitrogen) and purified with RNA purification beads e e Ampure XP beads (Beckman Coulter). Libraries were prepared starting from 100 ng of RNA per sample with TruSeq Stranded Total RNA Sample Prep Kit (Illumina) and sequenced with an Illumina NovaSeq 6000 sequencer, paired-end, covering 35 million reads. RNA-Seq data were aligned with Salmon (GRCh38) and processed as follows. DEG list generation: analysis was performed using edgeR (Section 3.24.3). Significant DEGs were filtered with FDR threshold at 0.05 and log FC threshold at 1.5. Gene comparison between different cell types: counts were normalized with the function cpm() of edgeR, with parameter log=T. Data were then visualized in a heatmap generated with the function heatmap.2() (gplots v.3.0.1.2).

[0125] NanoString analysis

[0126] hCD45 + CD3 + Cells were sorted from the spleen of two NSG nude mice at 10 and 18 wpt (respectively 8081 and 1869 sorted), lysed and hybridized to a Chimeric Antigen Receptor-T cell panel for characterization of 780 human genes (NanoString Technologies) overnight at 65°C. Hybridized samples were processed on the Prep Station and data were collected on the Digital Analyzer (NanoString) following the manufacturer’s instructions. Hybridized samples were processed on the Prep Station and data were collected on the Digital Analyzer (NanoString) following the manufacturer’s instructions. Raw data were imported into nSolver 4.0 (NanoString) for data quality check, background thresholding and normalization. The presence of 6 spiked-in RNA positive controls and 8 negative controls in the panel was used to confirm the quality of the run. Background level was determined by the mean count of the negative control probes and 2 x s.d. Samples containing probes with less than 50% above background or with imaging or positive control linearity flags were excluded from further analysis. Probes with raw counts below background in all samples were excluded from differential expression analysis to avoid false positive results. Data were normalized by the geometric mean of 10 housekeeping genes present in the CAR-T characterization panel.

[0127] Histology

[0128] Human thymus tissue and scaffold samples were fixed in 4% PFA (2 hours to overnight) and processed for either frozen or paraffin embedding. For frozen embedding, fixed tissues were equilibrated in 25% sucrose and embedded in O.C.T. compound (VWR). Frozen sections (7 pm thickness) were cut on a Leica cryostat 3050. For paraffin embedding, a Leica Peloris II tissue dehydrator and Sakura tissue Tech embedding station were used. Paraffin sections (3-5 pm thickness) were made using a ThermoFisher rotary microtome. Frozen or paraffin sections were stained with haematoxylin-eosin using an automatic station (Tissue Tek Prisma), a Masson’s trichrome kit (Leica, Raymond A Lamb, BDH Chemicals) and a Van Gieson staining kit (Millipore Merck).

[0129] Immunostaining

[0130] Tissue sections or coverslips were fixed in 4% PFA while blocking and permeabilizing with 5% normal donkey serum (NDS, Jackson Immuno Research) in PBS solution containing 0.5% Triton X (Triton™ X-100, SIGMA-ALDRICH). Tissue sections or cells were incubated with primary antibody 5% NDS, 0.01% Triton™ X solution overnight at 4°C. Secondary antibody was incubated for 45 minutes at room temperature (RT). Cell nuclei were counterstained with Hoechst 33342 (10"6M) or DAPI in Fluoroshield mounting medium (Abeam). TM

[0131] Imaging

[0132] ​Phase contrast images of cultured cells were acquired using an Olympus CK40 inverted microscope and an Olympus SC50 camera. Real-time imaging of cultured cells was acquired by a microscope Zeiss Axiovert 135 using a Hamamatsu Orca R2 camera (objective 10x NA 0.25 Plan Neofluar). For histological images, a Zeiss Axioplan2 microscope and a Zeiss Axiocam HRc color camera were used. Gross microscopy imaging of organs was performed using a Zeiss StREO Discovery.V20 microscope and a Zeiss Axiocam 506 color camera. Zeiss LSM710 inverted confocal microscope and Zen-Black software were used to acquire immunofluorescence images. Confocal images were processed using Fiji and Improvision Volocity LE software.

[0133] Vascular microsurgery, perfusion and decellularization

[0134] Male rats with body weight between 150-220 g were used as source of thymic organs. The thymus lacks a common artery that supplies blood to the whole organ, so it is not possible to perfuse the thymus by inserting a cannula in the main blood vessel as it is possible for other organs 35 . Microsurgical methods were employed to obtain thymic whole organ perfusion. Briefly, the blood vessels supplying each lobe were closed with silk sutures (size 6.0, F.S.T.) to occlude the thymus in the following order: right common carotid artery, right subclavian artery, right internal mammary artery, right costocervical trunk, right aortic arch, left aortic arch, left costocervical trunk and left internal mammary artery. The left carotid artery was left open for subsequent cannulation and organ perfusion. Cannulated thymuses were decellularized by perfusion of a detergent combined with enzymatic digestion (DET). The organ was perfused with dH20 (18.2 mW / cm) at 4°C (0.2 ml / min) for 96 hours using an i150n peristaltic pump (i Pumps). DET was performed at room temperature (RT) with 4% sodium deoxycholate (SDC; SIGMA-ALDRICH) and 0.1 mg / mL DNase-I (SIGMA-ALDPRICH) in 2.5 mM MgCl2(SIGMA-ALDRICH). Decellularized rat thymus (scaffolds) were gamma-irradiated with a dose of 1780 Gy and stored in PBS at 4°C for several weeks.

[0135] Micro-CT

[0136] Rat thymuses were extracted, cannulated and fixed in paraformaldehyde and then immersed in a 1 : 1 ratio of potassium triiodide (I2KI) to formalin solution injected into the cannula and then the rat thymuses were immersed into a centrifuge tube containing the same solution (approximately 20ml volume). I2KI was used with a total iodine content of 63.25mg / mL (iodine mass of 2.49 x 10 -4 mol / mL). The samples were immersed in the contrast solution for 48 hours before scanning. Specimens were rinsed in distilled water to remove excess surface iodine and dried with a gauze pad before micro-CT examination. They were wrapped in M and fixed in low-density plastic cylinders. The isotropic voxel size varied according to the geometric magnification factor achieved (inversely proportional to the sample size), ranging between 5 and 9 pm. X-ray images were acquired using an XT H 320 microfocus CT scanner (Nikon Metrology, Tring, UK) with a multi-metal target. A tungsten target was used with an accelerating voltage of 100 kV and a current of 100 microamperes. After micro-CT examination (duration of approximately 90 minutes), the samples were fixed in 10% formalin to prevent tissue degradation and to aid in the removal of iodine prior to macroscopic examination. Images were reconstructed using CT Pro 3D (Nikon Metrology, Tring, UK) and post-processed using VGStudio MAX v 3.0 (Heidelberg, Germany).

[0137] Scanning Electron Microscopy (SEM)

[0138] Samples were fixed in 4% formaldehyde / 2.5% glutaraldehyde (SIGMA-ALDRICH) in 0.1 M phosphate buffer (PB) at pH 7.4 for 5 hours at room temperature. Samples were then washed in 0.1 M PB, cut into approximately 5 mm and cryoprotected in 0.05 M PB containing 25% sucrose and 10% glycerol overnight. Samples were then flash frozen and fractured in liquid nitrogen using a stainless steel probe and then thawed back in cryoprotectant at room temperature. Samples were stained in 1% Os04 / 1.5% Potassium Ferricyanide, washed in H20 and dehydrated in a graded ethanol series, critical point dried using CO2 using a Leica EM CPD300 and mounted on aluminium stubs using adhesive carbon tabs. Mounted samples were coated with a thin layer of platinum using a Quorum Q150R Sputter Coater. Macro images of samples were taken prior to SEM imaging on a Leica M205C stereomicroscope. SEM images were recorded using a FEI Quanta 250FEG Scanning Electron Microscope.

[0139] DNA quantification

[0140] DNA was extracted from natural rat thymus and decellularized organs using the PureLink Genomic DNA Mini Kit (Invitrogen) according to the manufacturer's instructions. DNA samples were measured spectrophotometrically using NanoDrop. TM 1000, Thermo Fisher).

[0141] Thymus scaffold refilling and in vitro culture

[0142] Using an insulin syringe (Terumo, 29.5G), a cell suspension (2-3M cells, 100 μl / leaf) of TEC and TIC cells cultured in cFAD medium at a ratio of 5:1 was injected into decellularized rat thymus. The stromal cells used for scaffold reproliferation were derived from 4-month-old and 6-month-old donors. The seeded thymus scaffolds were cultured in cFAD medium for 5 or 6 days, and then in 50 μl of co-culture medium (DMEM 1X (Gibco), 10% FBS (SIGMA-ALDRICH), 1% penicillin and streptomycin (100X, SIGMA-ALDRICH), triiodothyronine (T3) (2 x 10⁻⁶)) (2 x 10⁻⁶ cells / leaf). -9 M SIGMA-ALDRICH, insulin (5 μg / ml, SIGMA-ALDRICH), and cytokines (interleukin-7, 5 ng / ml (Invitrogen), stem cell factor 5 ng / ml (Cell Signalling), and FLT3-L (5 ng / ml, CellGS)). Scaffolds were reconstructed using TEC alone (n=4), TIC alone (n=2), or TEC and TIC (n=5). Fully reconstructed scaffolds (using TEC, TIC, and TN) were preserved in co-culture medium (without cytokines, n=3) and cultured in cFAD medium for 14 days.

[0143] Bone marrow reconstruction and subcutaneous transplantation

[0144] NSG (8 to 12 weeks old) mice 137 Sublethal irradiation was administered at 3.75 Gy from a cesium source (IBL637 γ-radiator). For initial implantation, each mouse was intravenously injected with purified CD34. + Cells (100KCB-CD34) + Or 200KFL-CD34 + / cells). Evaluation of human cell transplantation in mouse bone marrow at death. In CD34 +Scaffolds were implanted subcutaneously in NSG mice 4 to 6 weeks post-injection. Mice were anesthetized with a 5-2% isoflurane-oxygen combination for induction and maintenance. Buprenorphine 0.1 mg Kg-1was given at induction for analgesia. Under sterile conditions, 1-3 midline incisions (0.4 cm) were made on the back of the mice and scaffolds were inserted in lateral pockets 40 Mice were sorted at different time points from 1 week post-transplantation (wpt) to 22 wpt.

[0145] In three independent experiments, a total of 16 NSG mice were subcutaneously transplanted, 14 of which were humanized with CB-CD34 cells and 2 with FL-CD34 genes: all mice showed bone marrow reconstitution. Mice were engrafted with 37 refilled scaffolds, of which 34 were retrieved. Specifically, 8 mice were engrafted with refilled scaffolds containing TIC, TEC, VeraVec and CD34 + cells (n=18); 6 mice received refilled scaffolds without CD34 + cells (n=13), and another 2 mice received scaffolds without VeraVec cells (n=6). In addition, two mice received empty scaffolds (n=2) and two mice received scaffolds with TIC only (n=6). The absence of endothelial cells in the scaffolds was not associated with a significant change in the vascular extent.

[0146] NSG nude mice (8-12 weeks old) were sublethally irradiated with 3.25 Gy and subsequently intravenously injected with purified human CD34 + cells. The humanization rate of NSG nude mice was lower than that of hairy littermates (4 out of 12 mice showed 27-55% hCD45 + cells, while 5 out of 5 hairy littermates showed 50-70% hCD45 + cells) as determined by bone marrow (BM) transplantation. Mice without BM reconstitution were excluded from further analysis. Bone marrow and spleen were retrieved from humanized mice engrafted with matrix refilled scaffolds (one at 10 wpt, one at 18 wpt) and subjected to flow cytometry analysis or sorting. Two humanized NSG-nude mice engrafted with empty scaffolds did not show peripheral refilling at 18 wpt and 20 wpt (no hCD3 + cells were detected in the spleen).

[0147] Scaffold FACS analysis and sorting

[0148] Collagenase I (2 mg / mL, SIGMA, C0130), dispase II (1 U / mL, Roche) and DNAse I (80 g / mL, Roche) in RPMI1640 and 2% FBS were used. At the end of digestion, cell suspensions were passed through a cell strainer (100 pm) and then stained with antibodies for FACS sorting or analysis. A total of 16 scaffolds were sorted and / or analyzed, of which 9 scaffolds were used with CD34 + HSC repopulation, 7 scaffolds not used with CD34 + HSC repopulation. Four control scaffolds not repopulated and two scaffolds containing only TIC, CD34 + and VeraVec were harvested at 22wpt and 11wpt, respectively. Scaffolds were analyzed and no thymocytes were detected. We found CD4 / CD8 SP and DP in 14 repopulated scaffolds: in 9 scaffolds repopulated with CD34 + HSC, 7 of which were not used with CD34 + HSC, also in 7 scaffolds not repopulated with CD34

[0149] Statistical analysis

[0150] All experiments were performed in biological triplicate or more; the exact sample size (n) for each experimental group / condition is indicated as discrete numbers. Statistical analysis was performed using two-way ANOVA non-parametric unless otherwise stated. Graphs were generated using GraphPad PRISM 8.

[0151] The above embodiments are described by way of example only. Numerous changes and modifications can be made without departing from the scope of the application as defined in the appended claims.

[0152] Examples

[0153] Long-term in vitro expansion to identify clonal TECs

[0154] Human thymuses were obtained from 33 patients undergoing open-heart cardiac surgery at ages between 3 days and 11 years. Thymic tissue was enzymatically digested into single cell suspensions and plated on feeder layers after lethal irradiation to derive thymic epithelial cells (TECs). All thymuses, irrespective of donor age, contained clonal TECs that generated colonies that could be extensively expanded after weekly passaging. Subsequently, it was determined whether clonal TECs originated from the cortex and / or medulla. Cell sorting strategies were used to anatomically and prospectively isolate thymic stromal compartments, including epithelial (cortical and medullary TECs) and mesenchymal (TICs). After several cycles of enrichment, CD45 -The stromal population was increased from 0.02-2% (freshly dissociated tissue) to about 40-80% so that subsequent sorting could be performed on the basis of additional surface markers. EpCAM high CD205 medullary TECs (mTECs) versus EpCAM low CD205 + cortical TECs (cTECs). Unexpectedly, the mesenchymal marker CD90 (Thyl) was also expressed by a large proportion (>60%) of mTECs and almost all cTECs as well as TICs (as shown in Figure 1A CD49f, a pi6-integrin, contributes to the formation of hemidesmosomes and is expressed by cells of the basal layer of stratified epithelia, was differentially expressed in both mTEC and cTEC populations (as shown in Figure 1B CD49f + mTECs or cTECs ("type 1") and CD49f - mTECs or cTECs ("type 2") subpopulations.

[0155] RT-qPCR was used to assess the expression of established functional markers of newly isolated medullary (e.g. AIRE1) and cortical (e.g. beta-5T, CD205) epithelial cells, respectively. This resulted in differential expression of cytokines and key transcription factors specific for the thymus and / or thymopoiesis (e.g. FOXN1 and PAX1) in cortical versus medullary cells, whereas expression of vimentin (VIM) was comparable in all tested classification populations.

[0156] To assess the colony-forming potential of newly isolated type 1 and type 2 human cTECs and mTECs, they were plated as described above with feeder cells after lethal irradiation. Colony- forming efficiency (CFE) assays consistently showed that the highest clonality existed in type 1 mTECs (CFE 2-4%) and type 1 cTECs (1-2%), whereas CFE of type 2 mTECs was very low (0.1-0.2%), and type 2 cTECs did not produce any expanded colonies (as shown in Figure 1CTECs, and the 1-type cTECs, although less abundant than the 1-type mTECs, expanded vigorously in culture and reached comparable numbers. To determine the transcriptional profile of clonal cells in the thymus, whole genome RNA sequencing was performed on cultured 1-type mTECs, 1-type cTECs and TECs cultured without any sorting strategy. The expression of genes detected by RNA sequencing of cultured TECs such as SIX1, EYA1, HOXA3, PAX1, PAX9 confirmed that clonal TECs maintained thymic identity independently of their origin zone (cortex or medulla). Both cortical (CSTV, KNICP3, CD274) and medullar (CHD1, EpCAM, CD24) genes were retained by cultured cells. When transcription factors of importance for thymic epithelial cells were analyzed by RT-qPCR, both mTEC and cTEC clonal cells expressed comparable levels of RNA encoding TRP63, FOXN1 and signaling molecules such as SCF. mTECs and cTECs showed only 11 differentially expressed genes (0.08%) out of 14347 genes, confirming that they displayed a common phenotype in culture. Therefore, in subsequent experiments, we referred to "clonal TECs" regardless of whether they originated from the medullar or cortical zone.

[0157] Single cell RNA sequencing defines a cell cluster common to medullar and cortical populations in vivo

[0158] Single-cell RNA sequencing (scRNA-seq) was performed on the newly isolated mTEC and cTEC subpopulations, type 1 and 2, to place them in the broader context of other mouse and human scRNA-seq studies. This combined approach provided an opportunity for high-resolution analysis of the newly isolated clonally enriched thymic epithelial cells. When the mTEC and cTEC subpopulations emerged in a single UMAP, the major cell subpopulations in the sorted epithelial cells could be annotated, defining 15 cell clusters that were distinguishable based on shared characteristics: cTEC groups in three major clusters, while mTEC groups in five major clusters, in addition to which we identified four TEC clusters that were shared by the cortex and medulla (“comTEC”). Note that the remaining three clusters (residual cells) represented polymorphonuclear leukocytes, dendritic cells, and thymic cell contaminants in the single-cell sorting. These major cell subpopulations were verified by specific gene markers for cTEC, mTEC, or comTEC. The comTEC clusters included cells characterized by epithelial-mesenchymal transition (EMT) features (cluster 12), proliferation markers (cluster 13), and cells expressing vimentin (cluster 14), while cluster 15 showed a unique signature enriched for ion transporters. When the transcriptional signatures of the comTEC and specific mTEC and cTEC clusters were compared to the transcriptional signatures of clonal TECs in vitro, it became clear that the latter expressed transcriptional signatures that were present in comTEC (proliferation, EMT, and vimentin), but not in other clusters of newly isolated mTEC or cTEC.

[0159] TECs co-express epithelial and mesenchymal markers, but differ from thymic mesenchymal cells

[0160] Based on the significant co-expression of CD90 (Thy-1) with EpCAM observed by FACS (as shown in Figure 1A and the detection of EMT signals in the comTEC clusters of newly isolated TECs, it was suggested that human thymic epithelial cells can be characterized by a unique epithelial-mesenchymal hybrid phenotype. Immunohistochemical analysis of postnatal thymus from our donor clearly detected cells that co-expressed cytokeratin (CK) with mesenchymal markers such as CD90, vimentin (VIM), and the mesodermal marker TE-7 (to a lesser extent), thus confirming the hybrid phenotype expressed in healthy thymus tissue at the protein level. In culture, clonal TECs were CD49f+, in active cell cycle (Ki67 +) and CK8 / 18 double positive. In contrast to all epithelial cell cultures described so far, but in agreement with the above reported, clonal TECs stably co-express mesenchymal markers such as CD90, VIM and TE-7. Moreover, some TECs expressing TRP63, a transcription factor of the basal layer of stratified epithelia and a major regulator of thymic epithelial cells, are also VIM positive in vitro and in vivo. Whole genome transcriptome analysis of cultured TECs confirmed the expression of EMT features defined by scRNA sequencing in vitro. Strikingly, the expression of mesenchymal markers is also associated with high motility and migratory properties typical of mesenchymal cells, as demonstrated by real-time imaging of TEC colonies cultured for 4 days. Thus, the hybrid epithelial-mesenchymal phenotype of epithelial cells appears to be a highly distinctive feature of the thymus in vivo and in vitro. In addition to TECs, CD90 + EpCAM - Mesenchymal cells also occur in expansion (as shown in Figure 1A These cells, not grown in culture conditions designed for epithelial cells, extensively proliferate for many passages in mesodermal culture conditions, thus allowing us to obtain billions of cells in a few weeks. Immunofluorescence shows the expression of mesenchymal markers such as TE7, VIM, PDGFRbeta, Chondroitin sulfate proteoglycan 4 (NG2) and smooth muscle actin (aSMA) (Figure 3d). FACS analysis shows that these TICs persistently express PDGFRalpha, PDGFRbeta, CD90, CD146, and to a different extent NG2 and alkaline phosphatase (ALP), in agreement with a perivascular phenotype. FACS analysis of expanded TICs shows the absence of epithelial cells (CD49f + ) or immune cells (hCD45 + ) Finally, we compared the transcriptional profile of cultured epithelial clonal cells and thymic mesenchymal cells in different age children samples by whole genome RNA sequencing. As expected, cultured TECs appear significantly different from TICs as shown by principal component analysis (PCAl). However, the data confirm that these two cell types share striking similarities in the expression of genes related to EMT, including high expression of COLlAl, MMP2, TGFBR2 and FNl.

[0161] Native thymic scaffold

[0162] Considering the remarkable ability of TECs and TICs to expand significantly under the culture conditions employed, it is possible to obtain billions of TECs and TICs in a relatively short time, making them putative stem / progenitor cells with clinical applicability. To assess their functional differentiation capacity, a unique approach was used to obtain the thymic extracellular matrix (ECM) by whole organ decellularization, which will provide a 3D structure for the seeding of thymic stroma and direct cell reorganization according to the physiological pattern.

[0163] Simply put, one carotid artery is kept open so that cannulation can be performed to allow organ perfusion, while all other arteries (downstream of the vessels that lead to the thymus) are closed, as shown under a gross microscope, as illustrated in the Method section below (Figure 2a, upper).

[0164] 3D visualization of a 24G cannula located in the carotid artery was achieved using X-ray micro-focused computed tomography (Micro-CT) of the entire rat thymus (Figure 2b). Micro-CT is a non-invasive technique that displays the blood-containing vascular system (white, dense area) via iodine contrast imaging; furthermore, it facilitates clear delineation of the cortex (brighter area, C) and medulla (M) regions through virtual segmentation, thus demonstrating high-resolution whole-organ 3D imaging of the thymus (Figure 2b). When the parenchyma tissue becomes more transparent, the 3D whole-organ micro-CT images enhance the 3D thymic vascularization segmented in red (Figure 2c). Standard histology, including Manson's trichrome (MT) staining and hematoxylin and eosin (H&E) staining, confirms the distribution of cortical and medullary regions and vascular structures (Figures 2d and 2e).

[0165] This microsurgical method involves injection via a single cannula. The compound, used to perfuse both thymic lobes, allowed for whole-organ imaging via micro-CT after coagulation and fixation (Figures 2f and 2g). Perfusion via a MilliQ-H20 cannula, followed by the addition of detergent and deoxyribonuclease solution, enabled whole-organ decellularization over 6 days (Figure 2a, lower). MT and H&E staining revealed complete organ decellularization, showcasing a fine, reticular, collagen-rich ECM with preserved macrovascular and microvascular systems (Figures 2h and 2i). The thymic ECM was characterized by colorimetric staining of ECM proteins, confirming the preservation of elastin and the absence of intermediate filaments, including keratin. Similarly, virtually no DNA remained after decellularization. Scanning electron microscopy (SEM) revealed the distribution of ECM fibers at high resolution, preserved in a gamma-radiation-sterilized whole-organ scaffold, which could be stored in physiological solutions.

[0166] Natural scaffolds promote in vitro morphogenesis of the functional thymus

[0167] The scaffolds were injected with 4 to 6 million clonal TECs, which had expanded for 4 to 6 weeks. The refilled scaffolds were then held in epithelial culture medium under static conditions for up to 14 days, during which time they showed scaffold remodeling and a gradual increase in volume (e.g., ...). Figure 3A (As shown). TEC extends along the 3DECM structure from the subcapsular region (CD49f). highreorganize to the innermost region of the scaffold. The seeded TECs express CK5 / 14 + and / or CK8 + ; are in active cell cycle (Ki67 + ); negative for the apoptosis marker Caspase3; and express TRP63 extensively (as shown in Figure 3B These data suggest that TECs are supported by 3DECM.

[0168] However, under these in vitro conditions, in the absence of any other cell type (as shown in Figure 3C), the seeded cells detected in patients affected by primary immunodeficiencies resemble TECs, with a severe disruption of the thymic stroma distribution.

[0169] To solve this problem, we next seeded expanded TECs with cultured TICs. To avoid excessive growth of mesenchymal cells, the ratio of TECs to TICs was optimized to 5:1. Remarkably, the presence of TICs promoted the organization of a cord-like stroma after only 5 days of co-culture, which largely replicated the early (9 weeks post-conception, wpc) fetal thymus (as shown in Figures 3D and 3E). TICs were found to be essential for the reorganization of TECs along the 3DECM of the reconstituted scaffold. In the presence of TICs, cortical CK8 + TECs were surrounded by a subcapsular layer of TECs expressing CD49f and TRP63 (as shown in Figure 3F Phenotype replication of the native thymus was found (as shown in + Scanning electron microscopy (SEM) of the re-populated scaffold confirmed cell-cell and cell-matrix interactions between TECs, TICs, and the acellular ECM, as well as the preservation of the vascular structure. The organized thymic stroma reconstituted in vitro by clonal cells was compared to clinical thymus sections cultured on sponges for up to 21 days to remove donor thymocytes, before transplantation into athymic patients 20 , 37 Considering the current clinical protocol for thymus section preparation, which shows a disorganized stroma, variable survival, and, importantly, residual CD3 + Thymocytes pose a significant risk of graft versus host disease to the transplant recipient, and the health of our reconstituted stroma in the scaffold appears particularly important.

[0170] To assess the functional potential of the 3D thymic microenvironment, as described, native scaffolds were re-populated with TECs and TICs and cultured for 4-6 days before injection of thymocyte precursors. Considering the size of the re-populated scaffolds (0.5-1 cm 3), we cultured cells for a relatively short time (up to 2 weeks), which is insufficient compared to the 5 to 6 weeks required for differentiation of the hematopoietic stem cell (HSC) lineage. To accommodate this, the three negative thymocytes (TN, CD3 - CD4 - CD8 - ) were re-populated with flow cytometry sorted three negative thymocytes (TN, CD3 Figure 3G CD4 + and CD8 + ) that are characterized by the expression of early lymphoid markers CD1a, CD5, CD7 and CD31 (as shown in Figure 3H ). The immune cells within the re-populated scaffolds were negative for CD33, positive for CD1a, and could differentiate towards double positive (DP) and mature single positive (SP) CD4 + : CD8 + cells, such that the ratio of CD4 + : CD8 + cells was similar to the phenotypic ratio in native human thymus (as shown in Figure 3H ). The EpCAM + TECs within the scaffolds also expressed HLA-DR, which was not detected on TECs expanded in culture. In contrast, TN injected into scaffolds containing only TIC stroma did not survive. In summary, organ size 3D structures could be cultured that support the survival and proper spatial organization of thymic stromal cells, thereby facilitating the differentiation of thymocytes.

[0171] Summary of human thymus morphogenesis in vivo

[0172] It was then determined whether the re-populated thymic scaffolds could mature and maintain HSC-repopulated human T cells in vivo. Scaffolds re-populated and cultured in vitro for 5 to 6 days were implanted subcutaneously in humanized NOD.scid. Il2Ryc null (NSG) mice. NSG mice were sub-lethally irradiated and transplanted with highly purified CD34 + HSCs from human cord blood (CBCD34 + ) or fetal liver (FLCD34 + ) 4-6 weeks prior to their implantation into thymic scaffolds co-seeded with stromal cells (TECs and TICs) and CD34 + HSCs at defined ratios. HSCs can be added to the transplant as a source of human myeloid cells to accelerate thymic development, but are not essential.

[0173] Decellularized native scaffolds pre-seeded and cultured in vitro were transplanted subcutaneously into NSG mice. At least 2 scaffolds were transplanted into each mouse. Grafts were harvested at 1, 2, 8, 11, 18 and 22 weeks post transplantation (wpt) respectively (as shown in Figure 4shown). Early time points were important to confirm survival and reorganization of stromal cells. At 8wpt, stromal cells seeded in native scaffolds (TEC + ) still showed a cord distribution or tight epithelial structure compared to the unique 3D thymic epithelial network (TEN) Figure 4 ). However, by 11wpt, ordered morphogenesis further developed, with more mature thymic structures consistently detected in scaffolds seeded with stromal and CD34 + HSCs. Among these were structures very similar to Hassall's bodies (HB), an anatomical feature of the human thymus, less pronounced in mice Figure 4 HBs indicated by asterisks.

[0174] Immunohistochemistry (IHC) for epithelial marker E-cadherin (ECad, red) and T cell receptor (TCR) (anti-CD3; green) showed thymic morphogenesis, with epithelial cells forming HBs and interacting with CD3 + cells on the scaffold. Epithelial (CK5 + ) and mesenchymal (Vimentin + CK5 - ) cells were detected in scaffolds transplants, as well as angiogenesis (detected by immunostaining for Endoglin), lower panel). In some experiments, VeraVec endothelial cells were seeded in addition to TECs and TICs, as they were reported to favor faster vascularization in reconstituted bone marrow microenvironments. However, thymic morphogenesis and T cell development were independent of the presence of human endothelial cells at seeding, likely due to mouse vascularization of the scaffolds.

[0175] To determine whether reconstituted scaffolds attracted circulating hematopoietic progenitor cells, morphogenesis was examined at 18-22 weeks post-transplantation (wpt) and found that most re-populated scaffolds had undergone thymic maturation Figure 4 ), including scaffolds that had not been injected with HSCs prior to transplantation. In addition, AIRE expression gradually increased from 11wpt to 22wpt, with scattered CD11c+ dendritic cells (DCs) within several regions, with or without HBs, and expression of HLA-DR on DCs detected by IHC and IHS on TECs and FACS.

[0176] To functionally characterize the implanted scaffolds, grafts were harvested, dissociated into single cells, and analyzed by flow cytometry. CD3 + cells were a prevalent phenotype in the hCD45 + population in reconstituted thymic implants. hCD45 + CD3 +The high percentage of cells (30-80%) indicates that hematopoietic progenitors are influenced by the instructive thymic microenvironment provided by the cultured human stromal cells, as compared to the bone marrow (BM) of the same mice, which is mostly CD33 + myeloid and CD19 + B lineage cells repopulation.

[0177] Flow cytometry showed that scaffolds were able to support the development of thymic cells (77.5% of all cells engrafting the scaffold), SPCD4 + and CD8 + cells expressing TCRaP. Notably, the CD4:CD8 ratio of cells maturing in the scaffold favored CD4 + SP cells (about 3:1), a feature of human thymic cell development that was not seen when circulating human bone marrow-derived progenitors were allowed to differentiate in the endogenous thymus of NSG mice. Immature SP (ISP) and DP thymic cells were also present in the scaffold at different ratios, showing a sustained thymopoiesis, with a consistent representation of DP cells. Nonetheless, their presence was reduced compared to the reported direct ex vivo examination of human thymus, which can reasonably reflect the fact that further optimization is needed to establish the precise differentiation schedule and kinetics in a native organ. In contrast, at 11 wpt, or at 22 wpt, no CD3 + cells were recovered from scaffolds repopulated with TIC, VeraVec, HSC CD34 + (without TEC) or empty scaffolds.

[0178] hCD45 + CD3 + cells were sorted from thymic scaffolds and endogenous thymus of NSG mice at 18 wpt and then interrogated for their functional potential in vitro. CD3 / CD28 was used to induce proliferation of sorted cells for 12 to 28 days until sufficient cells were obtained for functional interrogation. Interestingly, the CD4 / CD8 ratio of scaffold-associated or endogenous thymus-associated cells was maintained upon culture expansion, suggesting that human and mouse thymic stroma have different "instructions" for this trait. Expanded CD4 + and CD8 + cells could produce IFNy and TNFa when stimulated in vitro with phorbol myristate acetate (PMA) and ionomycin (Io), while cells developed in human scaffolds produced IL2 to a much lesser extent than T cells derived from NSG thymus (Figure 6 hours). Notably, freshly isolated thymic cells from human thymus also responded to PMA - Ionomycin with low IL2 production.

[0179] Finally, the ability of the re-populated scaffolds to support peripheral T cell reconstitution was assayed in athymic NSG-Foxn1nu nμll (NSG nude) mice, where the scaffold was the only thymic stroma. NSG nude mice were humanized 4 to 8 weeks prior to transplantation. Littermate NSG (hairy) mice were used as positive controls for bone marrow (BM) and peripheral cell reconstitution (spleen). No T cells (hCD3 + ) were detected in the periphery of these mice when transplanted with empty scaffolds after BM reconstitution. In contrast, hCD3 + cells were detected in the periphery when NSG nude mice were transplanted with scaffolds previously re-populated in vitro with human thymic stroma. Notably, hCD45 + CD3 + cells were purified from the spleen of NSG nude mice at two time points (10wpt and 18wpt) and analyzed for expression of specific genes supporting TCR activation, including TRBC1 / 2, TRAV19, NT5E, CD45RA, TRAC, CD69, PTPN6 (SHP-1), PTPRC, PPIA, CD22, LTB.

[0180] Discussion

[0181] The above results show that the human thymus after birth contains epithelial (TEC) and interstitial cells (TIC) that can be expanded in vitro to clinically relevant numbers suitable for reconstituting a human functional thymus in vivo. This shows that it is possible to replicate the long-lived phenotype of the human thymus using only post-natal cultured cells.

[0182] The results show that both CD49f-expressing mTEC and cTEC are enriched for clonal TEC with the ability to be passaged in culture and significantly expanded in vitro. Interestingly, expanded mTEC and cTEC mostly lost the characteristics of the respective originating compartment, which seems to be in line with the existence of common progenitor cells in the thymic tissue. Unexpectedly, clonal TEC co-express genes and exhibit some typical behavior of mesenchymal cells, a hybrid phenotype that is considered a unique cell intrinsic feature of the thymic stroma, which is stably maintained over many passages in culture.

[0183] A unique, bona fide population of interstitial cells (TIC) can be extensively expanded in vitro and share some features with other interstitial stromal cells and pericytes. These cells are important in guiding TEC during morphogenesis, both ex vivo and upon transplantation in vivo, as morphogenesis is defective in their absence. Their role can not be only to support the development and regeneration of TEC, but also to directly modulate T cell development in concert with TEC.

[0184] 3D native ECM proved to be very important for thymic morphogenesis of cultured cells. The whole organ perfusion decellularization method described above allowed to preserve the delicate 3D ECM network important for both ex vivo morphogenesis and long-term reconstitution in vivo. Moreover, human cultured TECs developed Hassall's bodies, proving their ability to maintain a key species-specific feature of thymic progenitors differentiation in vivo, which was not reported before. Functional genes such as AIRE and HLA-DR were also downregulated in the thymic stroma of human SCID patients in vivo, suggesting their expression is dependent on a complex microenvironment, including cross-talk between stroma and thymocytes. Therefore, the observed significant re-expression of AIRE and HLA-DR proves the extent of reconstitution achieved in the method of the present application. Key to prove the functional capacity of TECs and TICs when expanded in vitro was their ability to attract human HSCs from the circulation, to support T cell development and to re-populate around athymic NSG nude mice. Species-specific stroma in thymic organoids is essential for its own function; results show that the physiological human CD4 + / CD8 + ratio was consistently met. Importantly, human T cells developed in vivo within the human scaffold responded differently to stimulation than cells developed in the mouse thymus, indicating the instructive capacity of the human stroma.

[0185] In summary, in vivo both hematopoietic and stromal compartments of human origin gave rise to a durable thymus. Such a system opens the possibility to address many immunology questions, from T cell development, positive and negative selection of specific MHC-I and II, to unconventional T cells (such as gd), to establishment and maintenance of tolerance. Future applications can include thymus transplantation therapy for primary immunodeficiencies, such as athymic DiGeorge syndrome and Foxn1 nμll (hairless) babies; tolerance control in patients with congenital diseases (such as autoimmune polyglandular syndrome - candidiasis - ectodermal dysplasia (APCED)), and organ transplantation without immunosuppression.

Claims

1. A method of producing a thymic construct suitable for implantation into a subject, the method comprising the steps of: (i) providing a decellularized scaffold; (ii) seeding the decellularized scaffold with thymic epithelial cells having proliferative and differentiative capacity; and (iii) culturing the seeded scaffold to produce the construct; wherein the thymic epithelial cells are CD49f+ and at least one of VIM+, TE-7+, and CD90+.

2. The method of claim 1, wherein the step (ii) comprises seeding the decellularized scaffold with thymic epithelial cells (TECs) and thymic interstitial cells (TICs).

3. The method of claim 2, wherein the TECs and TICs are seeded at a ratio of TEC:TIC of 3: 1 to 10:

1.

4. The method of claim 3, wherein the TECs and TICs are seeded at a ratio of TEC:TIC of 4: 1 to 8:

1.

5. The method of claim 4, wherein the TECs and TICs are seeded at a ratio of TEC:TIC of 5:

1.

6. The method of claim 2, wherein the thymic epithelial cells are medullary thymic epithelial cells (mTECs), cortical thymic epithelial cells (cTECs), or a combination of mTECs and cTECs.

7. The method of any one of claims 1-6, wherein the decellularized scaffold is produced by decellularizing an entire thymus or a portion of a thymus or a thymic lobe.

8. The method of claim 7, wherein the thymus, portion of a thymus, or thymic lobe is decellularized by perfusion with at least one decellularization media selected from the group consisting of a detergent, a protease, and a nuclease.

9. The method of claim 1, wherein step (iii) comprises injecting or perfusing thymic epithelial cells into and / or onto the scaffold.

10. A pharmaceutical composition comprising isolated thymic epithelial cells having proliferative and differentiative capacity and a pharmaceutically acceptable carrier; wherein the thymic epithelial cells are CD49f+ and at least one of VIM+, TE-7+, and CD90+.

11. The pharmaceutical composition of claim 10, further comprising isolated thymic interstitial cells.

12. The pharmaceutical composition of claim 11, wherein the thymic epithelial cells are medullary thymic epithelial cells (mTECs), cortical thymic epithelial cells (cTECs), or a combination of mTECs and cTECs.

13. The pharmaceutical composition of claim 11, wherein the TECs and TICs are seeded at a ratio of TEC:TIC of 3: 1 to 10:

1.

14. The pharmaceutical composition of claim 11, wherein the TECs and TICs are seeded at a ratio of TEC:TIC of 4: 1 to 8:

1.

15. The pharmaceutical composition of claim 12, wherein the TECs and TICs are seeded at a ratio of TEC:TIC of 5:

1.

16. A thymic construct obtained or obtainable by the method of any one of claims 1-9. ​ 17. Use of a thymic construct according to claim 16 for the preparation of a medicament for the treatment of a thymic, immune or autoimmune disease or a thymic congenital defect.

Images