Patent Application
20250171743
The invention relates to media for cultivating tooth organoids. The present invention further relates to tooth organoids and markers to characterise such organoids.
Teeth play essential roles in food mastication and speech. Moreover, tooth physiology is more and more highlighted to impact body health and disease. In contrast to the wealth of knowledge on tooth development, homeostatic maintenance and repair in rodents, tooth biology remains far from understood in humans. Although stem cells of the mesenchymal compartments such as dental pulp and periodontal ligament (PDL) have substantially been characterized, knowledge on human tooth epithelial stem cells regarding presence, phenotype and biological function is scarce [Yu & Klein (2020) Development 147, dev184754]. Some indications for their existence have been found in the Epithelial Cell Rests of Malassez (ERM), a network of epithelial cells that is present in the dental follicle (dental follicle) which encloses unerupted teeth and upon tooth eruption remains present in the periodontal ligament around the root [Davis (2018) J. Vet. Dent. 35, 290-298]. These nests of epithelial cells express some stem cell-associated markers, and may play a role in regeneration of enamel and periodontal ligament following injury and inflammation, although repair capacity appears limited in postnatal life [Davis (2018) J. Vet. Dent. 35, 290-298]. During tooth development, enamel is formed by epithelial cells called ameloblasts [Yu & Klein (2020) Development 147, dev184754]. It has been reported that epithelial cell rests of Malassez-derived cells, when co-cultured with dental pulp stem cells (DPSCs), can differentiate into ameloblast-like cells [Shinmura et al. (2008) J. Cell. Physiol. 217, 728-738]. However, 2D-cultured epithelial cell rests of Malassez show highly limited growth capacity and rapid loss of phenotype [Athanassiou-Papaefthymiou et al. (2015) J. Dent. Res. 94, 1591-1600; Kim et al. (2020) Int. J. Mol. Sci. 21, 1-16; Nam et al. (2014) Mol. Cells 37, 562-567; Nam et al. (2011) Mol. Cells 31, 355-36; Tsunematsu et al. (2016) Lab. Investig. 96, 1063-1075].
A powerful method to in vitro grow and expand tissue epithelial stem cells is provided by organoid technology. Although meanwhile derived from numerous organs, epithelial organoids have not been established yet from human tooth [Gao et al. (2021) J. Dental Res. 100, 454-463; Binder et al. (2020) Sci. Rep. 10, 4963]. A previous study reported that epithelial cell rests of Malassez, seeded in Matrigel, grew as ‘organoids’ [Athanassiou-Papaefthymiou et al. (2015) J. Dent. Res. 94, 1591-1600]. However, these structures were not deeply characterized, and did not adhere to the current hallmarks of tissue-derived organoids such as (clonal) derivation and expansion from epithelial tissue stem cells under WNT-promoting conditions, and robust and long-term expandability. Other studies described the construction of bioengineered 3D dental structures, however only from animal origin (mouse, rat, dog, pig) at embryonic or neonatal age and non-expandable [Nakao et al. (2007) Nat. Methods 4, 227-230; Duailibi et al. (2008) J. Dent. Res. 87, 745-750; Young et al. (2002) J. Dent. Res. 81, 695-700].
The present invention discloses the establishment of long-term expandable organoid cultures starting from human tooth (i.e. from the dental follicle of third molars). The organoids show epithelial stemness characteristics mirroring epithelial cell rests of Malassez stem cells, and display ameloblast differentiation property reinforced by the presence of TGFb or dental mesenchymal cells, thereby recapitulating epithelial cell rests of Malassez/dental epithelial stem cell (DESC) features and known in vivo processes. These organoid models provide a valuable research tool to explore human tooth epithelial stem cell biology and epithelium-mesenchyme interplay, at present only poorly understood, thereby paving the way to unravelling their roles in tooth homeostasis and potential repair. Moreover, the tractable biological tooth stem cell structures represent an appealing step toward dental regenerative replacement prospects.
Insight into human tooth epithelial stem cells and their biology is sparse. Tissue-derived organoid models typically replicate the tissue's epithelial stem cell compartment. The present invention discloses an epithelial organoid model starting from human tooth. Dental follicle (dental follicle) tissue, isolated from unerupted wisdom teeth, efficiently generated epithelial organoids that were long-term expandable. The organoids displayed a tooth epithelial stemness phenotype similar to the dental follicle's Epithelial Cell Rests of Malassez (ERM), a compartment containing dental epithelial stem cells. Single-cell transcriptomics reinforced this organoid-epithelial cell rests of Malassez congruence, and uncovered novel, mouse-mirroring stem cell features. Exposure of the organoids to epidermal growth factor induced transient proliferation and eventual epithelial-mesenchymal transition, highly mimicking events taking place in the epithelial cell rests of Malassez in vivo. Moreover, the epithelial cell rests of Malassez stemness organoids were able to unfold an ameloblast differentiation process, further enhanced by transforming growth factor-beta (TGFβ) and abrogated by TGFβ receptor inhibition, thereby reproducing TGFβ's known key position in amelogenesis. By creating a mesenchymal-epithelial composite organoid (assembloid) model, it is herein demonstrated that the presence of dental mesenchymal cells (i.e. pulp stem cells) triggered ameloblast differentiation in the epithelial stem cells, thus replicating the known importance of mesenchyme-epithelium interaction in tooth development and amelogenesis. Also here, differentiation was abrogated by TGFβ receptor inhibition. Novel organoid models are described, empowering the exploration of human tooth epithelial stem cell biology and function as well as their interplay with dental mesenchyme, all at present only poorly defined in humans. Moreover, the new models may pave the way to future tooth-regenerative perspectives.
The invention is further summarized in the following statements:
a Schematic of organoid culture set-up. Progressing development of organoid structures after seeding dissociated dental follicle (DF) in tooth organoid medium (TOM) (passage 0, P0), and robust passageability (brightfield pictures of indicated P). b Histological (H&E) and ultrastructural (TEM) analyses of tooth organoids grown in TOM for 14 days. Box and arrow indicate cuboidal epithelium (CE) and squamous epithelium (SE), respectively. c-e Brightfield phase-contrast images and immunofluorescence staining pictures for markers as indicated, of primary dental follicle tissue and full-grown (day-14) organoids. Arrows indicate double-positive cells of indicated markers. Boxed areas are enlarged. DAPI was used to label nuclei. Scale bars: 50 μm, unless indicated otherwise.
a Experimental overview of the scRNA-seq analysis. UMAP plot of the annotated cell clusters in the integrated dental follicle-organoid dataset. DF, dental follicle; ERM, Epithelial Cell Rests of Malassez; NK, natural killer cells. b Heatmap displaying the scaled expression of the top 10 differentially expressed genes (DEGs) per cluster. Genes specifically described in the text are highlighted in bold. c Dot plot displaying the percentage of cells (dot size) expressing indicated marker genes with average expression levels (colour intensity) (see scales) in the epithelial cell rests of Malassez and organoid clusters. d Indicated regulons projected on UMAP plot. The ERM cluster is magnified at the bottom. e Immunofluorescence staining for markers as indicated in primary dental follicle tissue and organoids at specified time points. DAPI (blue) was used to label nuclei. f Significant (FDR≤0.05) DEG-based GO terms enriched in epithelial cell rests of Malassez versus P1 organoids (top) or in P1 and P4 organoids together versus epithelial cell rests of Malassez (bottom). g Violin plots showing gene expression level of indicated stemness markers in P1 and P4 organoids. Immunofluorescence staining of P1 and P4 organoids for the indicated markers. DAPI (blue) was used to label nuclei. Scale bars: 50 μm.
a Timeline of experimental set-up (d, day). Immunofluorescence analysis and quantification of KI67+ cells (mean±SEM; n=3 biological replicates) in organoids cultured as indicated. DAPI (blue) was used to label nuclei. b Immunofluorescence staining for the indicated markers of full-grown organoids (day 14; P0) cultured in medium as denoted. Arrows indicate double VIM+CK5+ cells. DAPI (blue) was used to label nuclei. c Timeline of experimental set-up. Left part: brightfield pictures of organoid cultures (day 14) as indicated. Boxed area is enlarged. Encircled areas show cell growth at the bottom of the culture plate. Immunofluorescence staining of full-grown organoids (day 14; P5) cultured as indicated for the indicated markers. Right part: brightfield pictures and immunofluorescence (VIM) staining of cells grown at the bottom of the plate (day 14; P5). Boxed area is enlarged. DAPI (blue) was used to label nuclei. Asterisk mark for orientation. Scale bars: 50 μm, unless indicated otherwise.
a Timeline of experimental set-up (d, day). Immunofluorescence examination of ODAM in organoids from culture conditions and time points as indicated. DAPI (blue) was used to label nuclei. Corrected total organoid fluorescence (CTOF) quantification of ODAM in organoids at indicated time points (mean±SEM; n=3 biological replicates). b Gene expression pattern of ameloblast secretory- and maturation-stage markers in MIM-switched organoids at time points as indicated. Data are expressed as fold change relative to the organoids at switching to MIM (d0). Expression is normalized to expression of GAPDH. Data are mean of n=3 biological replicates. Right: Gene expression levels (relative to GAPDH) of AMTN in MIM-switched organoids at time points as indicated (mean±SEM; n=3 biological replicates). Below: Immunofluorescence staining for the indicated markers, and quantification of SOX2+ and P63+ cells in organoids cultured in MIM (mean±SEM; n=3 biological replicates). DAPI (blue) was used to label nuclei. c Immunofluorescence staining for the indicated markers in organoids cultured as specified. DAPI (blue) was used to label nuclei. d Alizarin Red S (ARS) staining of organoids cultured as specified. Arrows indicate ARS+ areas. Images below show negative control (i.e. hematoxylin only). Right: Ultrastructural (TEM) analysis of MIM-switched organoids. Boxed area is enlarged. Arrowheads indicate calcium phosphate crystals. Scale bars: 50 μm, unless indicated otherwise.
a Experimental overview of the scRNA-seq analysis. UMAP plot of the integrated dental follicle and organoid samples as indicated. ‘Primary’ means all dental follicle clusters. b Projection of indicated genes on the integrated UMAP plot. c Heatmap displaying the scaled expression of the top 10 DEGs per cluster in P4 versus P4-switch organoids. d Significant (FDR≤0.05) DEG-based GO terms enriched in P4-switch versus P4 organoids. e Indicated regulons (STAT2, MAF, FOXC2) projected on the integrated UMAP plot. Dot plot of predicted STAT2 or MAF regulon target genes in P4 and P4-switch organoids. Projection of TGFβI gene expression on the UMAP plot.
a Timeline of experimental set-up (d, day). Immunofluorescence examination for the indicated markers in organoids cultured as denoted. Boxed areas are enlarged. DAPI (blue) was used to label the nuclei. CTOF quantification of indicated markers in organoids cultured as specified (mean±SEM; n=3 biological replicates). b-c Gene expression levels (relative to GAPDH) of indicated markers in organoids cultured as denoted (mean±SEM; n>3 biological replicates). d Timeline of experimental set-up. Histological (H&E) analysis and immunofluorescence examination for the indicated markers in assembloids cultured as indicated. DAPI (blue) was used to label the nuclei. Dotted area demarcates the (VIM+) mesenchymal cells. Scale bars: 50 μm.
a Brightfield images of the development of organoid structures (P0; day 14) after seeding dissociated dental follicle (DF) in the medium as indicated (see text). b Organoids growing out from dental follicle-derived cell clusters (top) or from single cells (bottom) in tooth organoid medium (TOM; passage 0, P0; d, day). c Histological (H&E) analysis of dental follicle. Boxed area is enlarged. Arrows indicate Epithelial Cell Rests of Malassez (ERM). d Brightfield and epifluorescence pictures of eGFP+ and eGFP− cells generated from dissociated organoids and cultured as mixture in TOM (P6; day 0 and day 14). Arrow indicates a fluorescent (eGFP+) organoid and arrowheads point to non-fluorescent (eGFP−) organoids. e Brightfield image of organoid culture after seeding of the dissociated dental follicle tissue in TOM (P0; day 14), showing attachment of spindle-formed mesenchymal cells at the bottom of the culture plate (arrows), which are lost at passaging, being not present anymore in the first passage (P1; day 14) Consecutive magnifications are indicated by black and blue boxes, respectively. f Progressing organoids' development in TOM during a single-passage (P4) 14-day culture period showing brightfield pictures, organoid diameters (mean±SEM; n=3 biological replicates), and proportions of proliferative and apoptotic events as quantified through KI67 and cleaved caspase-3 (CC3) immunostaining, respectively (dots indicate biological replicates; n=2). g Proportion of apoptotic (CC3+) cells and diameters of day-14 organoids over different passages (mean±SEM; n=3 biological replicates). h Organoids derived from dental follicle of erupted wisdom teeth from patients as indicated. i Immunofluorescence staining for markers as indicated in primary dental follicle tissue and organoids (P0, day 14). DAPI was used to label nuclei. j Gene expression levels (relative to GAPDH) of indicated markers in dental follicle and organoids (P1) (mean±SEM; n=3 biological replicates). Scale bars: 50 μm, unless indicated otherwise.
a Dot plot displaying the percentage of cells (dot size) expressing indicated marker genes with average expression levels (colour intensity) (see scales) of the annotated cell clusters. UMAP representation of the distinct cell clusters, and UMAP plot of the different patients (Pat). b Violin plots showing the distribution of the number of genes detected per cell (nGene), the total unique molecular identifier counts per cell (nUMI) and the percentage of mitochondrial content (percent.mito) per sequenced sample as indicated. Dashed lines show cut-off values (see Methods). c Significant (FDR≤0.05) DEG-based GO term enriched in the lower-quality cell cluster based on the top 10 DEGs. Ultrastructural (TEM) analysis of full-grown organoids (P5; day 15). Boxed area is enlarged. Arrowhead indicates an apoptotic nucleus. d-e Projection of indicated genes on UMAP plot. epithelial cell rests of Malassez cluster is enlarged at the bottom. f Projection of ITGA6 expression on UMAP plot. epithelial cell rests of Malassez cluster is enlarged at the bottom. Brightfield pictures of organoid cultures from FACS-isolated ITGα6+ or ITGα6− cells in TOM (P0; day 17). Boxed area is enlarged. Arrows indicate attached spindle-formed mesenchymal cells in the ITGα6− cell culture at the bottom of the plate. g Projection of indicated genes on UMAP plot. epithelial cell rests of Malassez cluster is enlarged at the bottom. Immunofluorescence staining of dental follicle and (day-14) organoids for indicated markers. DAPI (blue) was used to label nuclei. h Violin plots displaying activity of indicated regulons in the epithelial cell rests of Malassez and organoid clusters. Scale bars: 50 μm, unless indicated otherwise.
a Projection of indicated genes on UMAP plot. b Timeline of experimental set-up (d, day) and brightfield pictures of organoid cultures (day 14) as indicated. Right: diameter of organoids in specified cultures (violin plot; n=3 biological replicates). c Immunofluorescence staining for indicated markers in organoids cultured in TOM+EGF (P0). Boxed area is enlarged. DAPI (blue) was used to label nuclei. Arrows indicate double P63+VIM+ cells. d Expression levels (relative to GAPDH) of indicated genes in organoids (day 14) cultured as denoted (mean±SEM; n=4 biological replicates). Scale bars: 50 μm.
a Immunofluorescence staining for AMELX in organoids cultured as denoted. DAPI (blue) was used to label nuclei. b Gene expression levels (relative to GAPDH) of indicated markers in organoids cultured in TOM at indicated time points (mean #SEM; n=3 biological replicates). c Timeline of in vivo experimental set-up (d, day). ARS and Masson's trichrome (TCM) staining of depositions in recovered hydroxyapatite scaffolds which had been seeded with organoids before subcutaneous implantation, or only with Matrigel (empty). Boxed areas are enlarged. Negative control of ARS involves hematoxylin staining only. Arrow indicates a still discernible organoid. d Timeline of experimental set-up (d, day). Immunofluorescence staining for indicated markers in full-grown organoids as specified. DAPI (blue) was used to label nuclei. Quantification of SOX2+ cells in organoids cultured as indicated (mean±SEM; n=3 biological replicates). Gene expression levels (relative to GAPDH) of indicated markers in organoids cultured as specified (mean±SEM; n=3 biological replicates). Scale bars: 50 μm, unless indicated otherwise.
a Projection of indicated genes on the integrated UMAP plot (see
a Gene expression analysis of indicated TGFβ pathway components in organoids cultured as denoted. Expression is normalized to expression of GAPDH. Data are mean±SEM of n=3 biological replicates. b Brightfield and fluorescent (eGFP) images of assembloids comprising g dental epithelial (organoid-derived) cells and mesenchymal cells (DPSCs; marked by eGFP), cultured in TOM+αMEM. c Time-line of experimental set-up (d, day). Immunofluorescence staining for indicated markers of organoids cultured as denoted. DAPI (blue) was used to label the nuclei. Scale bar: 200 μm, unless indicated otherwise. d-e Gene expression levels (relative to GAPDH) of indicated TGFβ pathway components in assembloids cultured in TOM+□MEM (mean±SEM; n=3 biological replicates).
Organoids are 3D cell constructs that self-develop by proliferative expansion from tissue's epithelial stem cells when the dissociated primary tissue sample (containing the stem cells as single cells or contained within cell clusters) is seeded into an extracellular matrix (ECM)-mimicking scaffold (typically, Matrigel) and cultured in a defined cocktail of growth factors replicating stem cell niche signalling (if known) and/or tissue embryogenesis. Among others, activation of wingless-type MMTV integration site (WNT) and epidermal growth factor (EGF) signalling are typically needed [Boretto et al. (2017) Dev. 144, 1775-1786; Cox et al. (2019) J. Endocrinol. 240, 287-3089; Sato et al. (2009) Nature 459, 262-265]. Resultant organoids duplicate the epithelial stem cell compartment of the tissue of origin in molecular phenotype and functional characteristics, and can generate differentiated tissue cell types under specified culture condition. As an important asset, organoid cultures can be serially expanded (passaged) without loss of characteristics, thereby providing a robust and faithful source of the primary tissue's epithelial stem cells and overcoming their generally limited availability and culture-ability. Typically, epithelial organoid models are established without the need for prior isolation of the epithelial (stem) cells from the dissociated whole-tissue sample since the accompanying mesenchymal cells do not thrive in the specific culture conditions used and are swiftly lost at culture and passaging.
The present invention reports the development of a long-term expandable epithelial organoid model derived from human dental tissue. The dental follicle-derived organoids show a stemness expression profile congruent with the epithelial cell rests of Malassez, previously advanced to encompass dental epithelial stem cells [Davis (2018) J. Vet. Dent. 35, 290-298]. In addition, single-cell transcriptomics uncovered novel molecular features (such as the stemness-associated hybrid E/M nature, new markers and gene-regulatory networks) for the as yet ill-defined and poorly comprehended human dental epithelial stem cells and epithelial cell rests of Malassez, often mirroring findings in mouse. Noticeably, organoid culturing appeared to proliferatively (re-) activate the stem cells of epithelial cell rests of Malassez, previously reported to be highly quiescent in vivo [Shinmura et al. (2008) J. Cell. Physiol. 217, 728-738]. Moreover, described (stem cell-related) functional properties of the epithelial cell rests of Malassez were markedly recapitulated by the tooth organoids. First, exposure to EGF induced transient proliferation and eventual EMT and migration, thereby mimicking events taking place in the epithelial cell rests of Malassez in vivo (for instance, upon tooth insult) [Davis (2018) J. Vet. Dent. 35, 290-298]. Second, the tooth organoids displayed the capacity to unfold an ameloblast differentiation process, as occurring in vivo during tooth formation [Yu & Klein (2020) Development 147, dev184754] and reported for epithelial cell rests of Malassez [Hamamoto et al. (1996) Oral Surg. Oral Med. Oral Pathol. Oral Radiol. Endod., 81, 703-709], thus recapitulating the epithelial cell rests of Malassez differentiation capacity. The organoids displayed molecular changes constituting pathways that underlie ameloblast differentiation during amelogenesis [Liu et al., (2015) BMC Genomics 16, 592; Nurbaeva et al. (2017) J. Physiol. 595, 3015-3039]. In addition, the organoids recovered the key position of TGFβ in ameloblast differentiation/amelogenesis [Benedete (2008) Pediatr. Dev. Pathol. 11, 206-212], as well as in periodontal ligament development [Davis (2018) J. Vet. Dent. 35, 290-Moreover, the present scRNA-seq interrogation advanced molecular transitions not revealed before in human amelogenesis. Also, STRING analysis projected protein-protein interactions that may further deepen knowledge on amelogenesis in human tooth, at present not understood. Together, the present model has the potential to in detail decipher ameloblast development and their production of enamel, the quintessential component of teeth, which would represent a leap forward in the dental field (especially for future dental tissue replacement therapies). Third, the organoid transcriptome reflected functional processes before (provisionally) assigned to the epithelial cell rests of Malassez, including regulation of bone mineralization, osteoblast differentiation and tooth eruption. Hence, the present model provides a tool to help decipher the multiple biological functions assigned to the epithelial cell rests of Malassez. The organoids show strong expandability, thereby overcoming current hurdles of primary epithelial cell rests of Malassez/dental epithelial stem cell culturing, such as limited cell number, life span and phenotypical loss. The expansion ability will be highly instrumental for allowing in-depth analysis of this yet enigmatic cell population. Finally, the induction of ameloblast differentiation by the presence of mesenchymal cells, thereby recapitulating the acknowledged importance of epithelium-mesenchyme interaction in tooth development including amelogenesis, again further corroborated the biomimetic value of the present model(s). Altogether, the several characterizations provide strong evidence that the present human tooth (dental follicle)-derived organoid models, present a valuable tool to study human tooth epithelial stem-cell biology and development, at present far from understood. Organoid technology is also highly applicable to human disease modelling in vitro. It has been suggested that epithelial cell rests of Malassez cells are associated with the pathogenesis of odontogenic cysts and tumours. Developing organoids from these lesions may help to gain better insight in their pathogenesis. More in general, the present tooth organoid approach can harnessed to model and study tooth diseases ranging from impact of bacteria to genetic mutations (like mutations in P63 and PITX2 associated with tooth anomalies and amelogenesis imperfecta), eventually leading to novel therapeutic targets and treatments.
Organoids have been shown amenable to regenerative replacement therapy. Damaged, lost or missing teeth, causing major health problems, may be regenerated or replaced by transplanting biological tooth constructs. Such approach may be superior (both material- and function-wise) to the traditional, still suboptimal synthetic implants, among others suffering from lack of physiological functionality, inferior bone integration and absence of innervation. Embryonically derived, bioengineered mouse tooth germs have been shown capable of forming a functional tooth unit after transplantation in an emptied dental cavity of the mouse. The present organoid and assembloid models may provide essential puzzle pieces toward developing human tooth germs. Although transplantation of natural teeth has been performed in some patients, especially children and young adolescents, the availability of such teeth remains limited. Of important note, the murine Matrigel should then be replaced by a clinically compatible ECM mimic. Currently, attempts are being made to substitute Matrigel for defined synthetic hydrogels, although achievements are still limited. In conclusion, a long-term expandable stemness organoid model from human tooth is developed, replicating molecular and functional features of the originating epithelial stem cell compartment. The new in vitro model will be highly valuable to explore human tooth epithelial stem cell phenotype and biology such as ameloblast differentiation. Moreover, the present invention indicates that the postnatal human tooth still contains epithelial stem cells, and the organoids will be beneficial to address the question on their role(s), and on the reasons why they do not, or not prominently, regenerate tooth tissue in postnatal life. This search also implicates the question whether these stem cells can in vivo be re-activated for repair. This understanding may eventually instigate tooth-regenerative approaches by re-activating endogenous repair capacity and processes.
The present invention discloses, as illustrated by Example 1, that epithelial organoids can be established from human tooth-derived dental follicle, displaying an epithelial cell rests of Malassez-mirroring, stemness expression phenotype and possessing robust long-term expandability.
The present invention discloses, as illustrated by Example 2, the present detailed scRNA-seq interrogation demonstrates and reinforces the organoid-epithelial cell rests of Malassez stemness relationship and uncovered new molecular fingerprints of human epithelial cell rests of Malassez, at present only poorly defined.
The present invention discloses, as illustrated by Example 3, that adding EGF to the organoids recapitulates functional in vivo behaviour of the epithelial cell rests of Malassez, thus advancing the present tooth organoid model as an interesting tool to study epithelial cell rests of Malassez phenotype and conduct, to date not well understood.
The present invention discloses, as illustrated by Example 4, that the herein disclosed tooth organoid model is capable of unfolding an ameloblast differentiation process involving known consecutive steps, thereby recapitulating dental epithelial stem cell/epithelial cell rests of Malassez functionality, and thus provides a valuable research tool to study amelogenesis of human tooth, at present poorly defined.
The present invention discloses, as illustrated by Example 5, single-cell transcriptomics of the tooth organoids driven into amelogenesis differentiation demonstrates and underscores the relevance of the present organoid model by confirming known data as well as presenting new insights in the amelogenesis process in humans which is at present far from clarified.
The present invention discloses, as illustrated by Example 6, that TGFβ coerces the tooth organoids into more pronounced ameloblast differentiation as well as into the direction of periodontal ligament development. These findings conform to the known activity of TGFβ in these tooth developmental processes, and thus again corroborate the strength and validity of the present organoid model. Moreover, they provide supportive evidence that the dental follicle-derived organoids replicate the multipotency of dental (HERS/epithelial cell rests of Malassez) epithelial stem cells as proposed to unfold in vivo during tooth development and possibly repair.
The present invention discloses, as illustrated by Example 7, that ameloblast differentiation of epithelial (organoid) stem cells is triggered by the presence of tooth mesenchymal cells involving TGFβ signalling, thereby corroborating in vivo findings of interactive mesenchyme-epithelium importance, and further validating this model as valuable research tool for exploring human tooth (stem cell) biology.
To develop epithelial organoids starting from human tooth, the dental follicle, known to encompass a large mesenchymal component but also the small epithelial epithelial cell rests of Malassez compartment, was isolated from unerupted third molars (wisdom teeth) extracted from adolescent patients (
aFor in vivo transplantation
bFor scRNA-seq analysi
Organoid structures gradually developed in two weeks time (
The developed organoid structures displayed a dense morphology (
To decode the organoids in more granular detail, single-cell RNA-sequencing (scRNA-seq) analysis w applied on dental follicle-derived organoids (at P1 and P4) together with their primary tissue (
Compiling the clusters' top 10 differentially expressed genes (DEGs) in a heatmap exposed specific expression patterns of the different clusters (
In further GO analysis, it was found that the biological terms ‘regulation of osteoblast differentiation’, ‘regulation of bone mineralization’ and ‘regulation of neuron projection development/neuron development’ are enriched in the epithelial cell rests of Malassez (
In general, gene expression signatures of P1 and P4 organoids display substantial similarity (
To establish organoids from primary tissues, supplementation of EGF is generally found quintessential. Hence, it is remarkable that dental follicle-derived organoids develop and expand in the absence of exogeneous EGF (TOM; Table 1). scRNA-seq mining exposed that the EGF receptor (EGFR) ligands amphiregulin (AREG) and heparin-binding EGF (HB-EGF), together with EGFR, are highly expressed in the organoids (
During tooth development, dental epithelial stem cells give rise to ameloblasts which produce enamel matrix proteins (EMPs) for amelogenesis [Yu & Klein (2020) Development 147, dev184754]. It has also been shown that epithelial cell rests of Malassez can differentiate into ameloblast (-like) cells [Hamamoto et al. (1996) Oral Surg. Oral Med. Oral Pathol. Oral Radiol. Endod. 81, 703-709] and produce EMPs. Ameloblast differentiation encompasses a secretory stage with production of the EMPs AMELX and ameloblastin (AMBN), and a maturation stage during which amelotin (AMTN) and odontogenic-ameloblast associated protein (ODAM) are produced. The EMPs are proteolytically cleaved by matrix metalloproteinase 20 (MMP20) and kallikrein (KLK4), typically expressed during the secretory and maturation phase, respectively. Here, it was examined whether the dental follicle-derived organoids, possessing an epithelial epithelial cell rests of Malassez-stemness phenotype, can be driven into differentiation toward ameloblasts.
Organoids expanded in TOM were switched to a medium previously reported to trigger ameloblast-like differentiation in 2D dental epithelial stem cell cultures [Yan et al. (2006) Eur. J. Oral Sci. 114, 154-158], (referred to as mineralization-inducing medium, MIM; Table 3), and analysed at multiple time points (
Interestingly, ODAM expression swiftly emerged (from day 2) in the organoids switched to MIM, and increased in intensity while remaining absent in TOM-cultured organoids (
Finally, to validate whether the ameloblast differentiation capacity is already present in the organoid-initiating epithelial cell rests of Malassez stem cells, the organoids were developed immediately in MIM and subsequently analysed their phenotype (in P1;
To decipher the amelogenesis differentiation process that occurs in the tooth organoids in deeper detail, scRNA-seq analysis was performed of P4 organoids switched to MIM for 8 days (referred to as P4-switch; see
As expected, stemness markers (e.g. SOX2, KRT15) are more prominent in the non-differentiated P4 organoid cluster, whereas ameloblast differentiation markers (e.g. AMTN, ODAM) show almost exclusive expression in the differentiated P4-switch organoids (
Looking more broadly at gene expression differences using DEG analysis revealed that P4 and P4-switch organoid clusters clearly display different gene signatures, thereby exposing interesting (new) markers (
GO analysis revealed enriched ‘negative regulation of cell differentiation’ in the straight P4 organoids when compared to P4-switch organoids (
Gene set enrichment analysis (GSEA) was performed which exposed several important differentiation (amelogenesis) pathways in P4-switch versus P4 organoids. Firstly, mineralization hallmarks (tooth, enamel) are significantly enriched in P4-switch organoids. In addition, calcium-signalling pathways, highly important during amelogenesis, were found significantly associated with the P4-switch organoids such as the hallmarks ‘calmodulin binding’, ‘store-operated calcium entry’ and ‘calcium mediated signalling’. Further interestingly, GSEA revealed significant enrichment of TGFβ signalling hallmarks in P4-switch versus P4 organoids, more specifically TGFβ (receptor) signalling and TGFβ (particularly TGFβ1/3) production, in line with the importance of the TGFβ pathway in amelogenesis [Benedete (2008) Pediatr. Dev. Pathol. 11, 206-212].
Regulon analysis exposed higher activity of the signal transducer and activator of transcription 2 (STAT2) gene-regulatory network in P4-switch than P4 organoids (
In a final analysis of the scRNA-seq data, STRING was applied to in silico predict protein-protein interactions. Using the top 40 DEGs in P4-switch versus P4 organoid clusters, it is projected that AMTN and ODAM closely interact (
The above analyses exposed the enrichment of TGFβ pathway processes in tooth organoids subjected to ameloblast differentiation, in line with the proposed key role of TGFβ in amelogenesis. To assess the impact of the TGFβ pathway, organoids grown in TOM (P5) were switch to MIM with or without TGFβ (
During tooth root development, Hertwig's epithelial root sheath (HERS), from which epithelial cell rests of Malassez is eventually derived, undergoes EMT to develop to participate in periodontal ligament formation. It has previously been proposed that this EMT process is regulated by TGFβ, thereby triggering HERS/epithelial cell rests of Malassez cells to switch phenotype toward periodontal ligament cells. Addition of TGFβ indeed further increased the expression of VIM in the epithelial (CK5+) organoids (
Given the importance of mesenchyme-epithelium interactions during tooth development including ameloblast differentiation/amelogenesis, it was investigated whether addition of dental mesenchymal cells had an impact on ameloblast differentiation of the epithelial organoids. It was chosen to use DPSCs to mimic early stages of tooth development in which DPSC-derived odontoblasts are in close contact with ameloblasts. The DPSCs, isolated, grown and characterized using well-defined standard protocols, were combined with organoid-derived epithelial stem cells in a layered approach, thereby forming composite organoids (assembloids) which were cultured in a mixture of TOM and the DPSC growth medium αMEM (
Whereas ODAM is not present in the straight (pure) epithelial organoids cultured in TOM (see above and
Third molars, predominantly unerupted, were extracted from adolescent patients (Table 2) at the ‘Oral and Maxillo-Facial Surgery-Imaging & Pathology (OMFS-IMPATH)’ unit of University Hospitals (UZ) Leuven after informed consent. The study was approved by the Ethics Committee Research UZ/KU Leuven (13/0104U). For sample collection, the gingiva was pushed aside after which the bone was perforated and the third molars with associated dental follicles were carefully isolated (without the visually distinct gingiva). dental follicle tissue was diligently peeled from the tooth and collected in Eagle's Minimum Essential Medium (αMEM; Sigma-Aldrich) supplemented with 10% foetal bovine serum (FBS; Sigma-Aldrich), 1% penicillin-streptomycin (Gibco) and 0.5% fungizone (Amphotericin B; Gibco). Following short rinsing steps in 70% ethanol and phosphate-buffered saline (PBS; Gibco), tissue was minced into small (˜1 mm2) fragments, and further dissociated using collagenase VI (3 mg/ml; Thermo Fisher Scientific) and dispase II (4 mg/ml; Sigma-Aldrich) for 2 h at 37° C., while regularly pipetting up and down. The single cells and few small cell clusters were collected through a 40 μm cell strainer (Corning) while removing the remaining larger and fibrous tissue fragments.
The dissociated dental follicle cell material was resuspended in a mixture of serum-free defined medium (SFDM; Thermo Fisher Scientific; Table 4) and growth factor-reduced Matrigel (Corning) in a 30:70 ratio, which was plated in 48-well plates at 20,000 cells per 20 μL drop. After solidification, tooth organoid medium (TOM; Table 1), unless indicated otherwise, was supplemented. ROCK inhibitor (RI; 10 μM; Merck Millipore) was added the first day of seeding (or passaging). Organoid cultures were kept at 37° C. in a 1.9% CO2 incubator, and medium was refreshed every 2 to 3 days, each time supplemented with fungizone (0.1%).
The organoid cultures were passaged every 10 to 14 days. Matrigel droplets were collected using ice-cold SFDM, and organoids dissociated using TrypLE (containing 5 μM RI; Thermo Fisher Scientific) and mechanical trituration. Remaining large organoid fragments were allowed to sediment and the supernatant, containing single cells and small fragments, seeded as described above. A split ratio of 1:6 was applied once the culture reached stable growth (typically from P2-P4). Organoids were cryopreserved and stored in liquid nitrogen.
To assess clonal derivation, dissociated single organoid cells were transduced with the lentiviral vector LV-eGFP during 30 min at 37° C., resulting in 60% eGFP+ cells as analysed by flow cytometry. The resulting mixture of eGFP+ and eGFP− cells was seeded in organoid culture as described above, and cultures analysed 14 days later using brightfield and epifluorescence microscopy (Axiovert 40 CFL; Zeiss).
FACS isolation of ITGα6+/− cells from dental follicle
Primary dental follicles were dissociated into single cells as described above. Cells were incubated with PE-anti-ITGα6 antibody (1:5; Cat.no 555736; BD Biosciences) and rinsed, both performed in TOM supplemented with fungizone (0.1%) and RI (10 μM). ITGα6+ and ITGα6− cells within the living (DAPI-negative) population were sorted in TOM (supplemented with fungizone and RI) using a BD Influx (BD Biosciences), and seeded at 7,500 cells per 20 μL Matrigel droplet as mentioned above. RI (10 UM) was added to the cultures for 1 week.
Organoids (or dissociated dental follicle) were cultured in mineralization-inducing medium (MIM; Table 3; time schedule, see
Matrigel (10 μl) with dissociated organoid cells (150,000) was pipetted into custom-made 3D-printed hydroxyapatite constructs (Sirris) which were subcutaneously transplanted in immunodeficient nu/nu mice (Janvier Labs), as in detail described in [Bronckaers et al. (2021) Methods Mol. Biol. 2206, 223-232.]. After 4 weeks, implants were resected and subjected to 48h-fixation in 4% paraformaldehyde (PFA) (Sigma-Aldrich), paraffin-embedding, 24h-decalcification, 7-μm sectioning and Alizarin Red S (ARS) or Masson's Trichrome (TCM) staining as described in Bronckaers et al. (2021) Methods in Mol. Biol. 2206, 223-232. The study was approved by the Ethical Committee on Animal Experiments (ECAE) of Hasselt University (protocol 202044).
DPSCs were obtained as in detail described and characterized in About et al. (2000) Am. J. Pathol. 157, 287-295. In short, dental pulp was collected from the extracted wisdom teeth (after careful removal of the apical papilla), minced and fragments cultured in T25 flasks (Corning) in αMEM supplemented with 10% foetal bovine serum (FBS) and 1% L-glutamine (Gibco). When 70-80% confluence was reached, cells were trypsinized and re-plated at 150,000 cells per T75 flask, and used at early passage (˜P5) for assembloid creation. For GFP labelling, DPSCs were transduced with the lentiviral vector LV-eGFP as described above.
Organoid and DPSC cultures were dissociated into single cells, and mixed in a round-bottom low-attachment plate (96-well; Greiner) using a layered approach [Nakao et al. (2007) Nat. Methods 4, 227-230]. First, DPSCs (5×104 cells) were sedimented by centrifugation (300 g for 1 min at 4° C.), followed by deposition of the organoid-derived cells (1×105; at 300 g and 4° C. for 2 min). The cells were layered in 10% Matrigel and 90% of a 1:1 mixture of TOM (i.e. organoid growth medium) and αMEM (i.e. DPSC growth medium), and then incubated for 24 h at 37° C. in 5% CO2. The formed aggregate was re-plated into a 48-well plate in a 20 μL Matrigel (70%) droplet as described above to generate the assembloid, further cultured in TOM+αMEM with or without the TGFβ receptor inhibitor LY2109761 (5 μM) as indicated.
Primary dental follicle tissue, organoids and assembloid were fixed in 4% PFA for 1h, embedded in paraffin, and sections subjected to hematoxylin and eosin (H&E), immunofluorescence or ARS staining. Antigen retrieval (10 mM citrate, pH6) and permeabilization (0.1% Triton X-100; Sigma-Aldrich) were performed. After incubation with primary and secondary antibodies (Table 5), sections were mounted with Vectashield (DAPI; Vector Laboratories) or DPX mountant (Sigma-Aldrich). Analysis was done using a Leica DM5500 epifluorescence microscope or a Zeiss Axioimager epifluorescence microscope. ImageJ software was used to quantify immunoreactive signal intensity and the ‘corrected total organoid fluorescence’ (CTOF) (=integrated density−(area of organoid×mean fluorescence of background readings).
Organoid samples were prepared for transmission electron microscopy (TEM). In short, samples were fixed in glutaraldehyde/osmium tetroxide, dehydrated, embedded in epoxy resin, and cut into 40-70 nm sections. TEM analysis was performed with the JEM1400 transmission electron microscope (JEOL) equipped with an Olympus SIS Quesmesa 11 Mpxl camera, or the Philips EM208 S electron microscope (Philips) equipped with the Morada Soft Imaging System camera with corresponding iTEM-FEI software (Olympus SIS).
RNA was extracted from dissociated dental follicle, organoids and assembloids using the GenElute Mammalian Total RNA Miniprep Kit (Sigma-Aldrich) according to the manufacturer's instructions. RNA was reverse-transcribed (RT) using the Superscript III First-Strand Synthesis Supermix (Thermo Fisher Scientific) and the resultant cDNA samples were analysed with SYBR Green-based quantitative PCR (qPCR) using the StepOnePlus Real-Time PCR System (AB Applied Biosystems). Forward and reverse primers (Table 6) were designed using PrimerBank and PrimerBlast. Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was included as housekeeping gene. Relative gene expression levels were calculated as ΔCt (Ct target−Ct housekeeping gene) and compared to control (see figure legends).
Primary dental follicle tissue from two individual patients (see Table 2), and derived organoids at P1 and P4, or switched to MIM (P4-switch), were dissociated into single cells (as described above) and subjected to scRNA-seq analyses using 10× Genomics, according to manufacturer instructions. Libraries were generated using the Chromium Single Cell 3′ v2 Chemistry Library Kit, Gel Bead & Multiplex Kit (10× Genomics), and sequenced on NovaSeq6000. After quality control, raw sequencing reads were demultiplexed, aligned to the human reference genome GRCh38 and processed to a matrix representing the UMI's per cell barcode per gene using CellRanger (v3; 10× Genomics). Downstream analysis was performed in R (v.3.6.1) using Seurat (v.3.0).
First, data from the primary dental follicle tissue, P1 and P4 organoids were integrated and analysed, and subsequently data from P4-switch organoids were added for a next analysis (further referred to as Integration 1 and Integration 2, respectively). Dead cells and potential doublets (i.e. with <300 genes or >10,000 genes, >150,000 unique molecular identifiers (UMI) and >15% mitochondrial RNA) were removed (
Gene ontology analysis (GO) of biological processes was done in Panther using significant differentially expressed genes (DEGs; FDR≤0.05 and log FC≥0.25). Gene-regulatory networks (regulons) were identified using SCENIC (pySCENIC; v.0.9.15) in Python (v.3.6.9). In short, co-expression modules were generated and regulons inferred (with parameters and hg38__refseq-default r80_10 kb_up_and_down_tss.mc9nr.feather and hg38__refseq-r80_500 bp_up_and_100 bp_down_tss.mc9nr.feather motif collections) resulting in a matrix of AUCell values that represent the activity of each regulon in each cell. The AUCell matrix was imported into Seurat and regulons were projected on the integrated UMAP plot.
Gene-set enrichment analysis (GSEA; v.4.1.0) was performed on P4 and P4-switch organoids using normalized expression data. Gene sets (hallmarks) tested were obtained from the Molecular Signatures Database (MSigDB; v.7.2).
To predict protein-protein interactions with STRING (v.11.0), the top 40 DEGs of P4-switch organoids versus P4 organoids were used. The cluster analysis was subdivided in three colours by kmeans. The minimum required interaction score was set as medium confidence (0.4).
Finally, the pseudotime trajectory was projected onto the integrated UMAP dimensional reduction generated previously with Seurat (P1, P4 and P4-switch organoids) using the Monocle3 (v1.0.0) package's learn_graph and plot_cells functions.
Raw sequencing data are available at ArrayExpress (accession number E-MTAB-10596).
Statistical analysis was performed using GraphPad Prism (v.9.0.0). (Un-) paired two-tailed t-student test was applied for comparison of 2 groups or two-way analysis of variance (ANOVA) for multiple comparisons followed by Sidak's test for Multiple Comparison. Statistical significance was defined as P≤0.05.
| Number | Date | Country | Kind |
|---|---|---|---|
| 22158089.7 | Feb 2022 | EP | regional |
| Filing Document | Filing Date | Country | Kind |
|---|---|---|---|
| PCT/EP2023/054422 | 2/22/2023 | WO |
