The skin is a multilayered organ, equipped with appendages (that is, follicles and glands), that is critical for regulating body temperature and the retention of bodily fluids, guarding against external stresses and mediating the sensation of touch and pain1,2. Reconstructing appendage-bearing skin in cultures and in bioengineered grafts is a biomedical challenge that has yet to be met3,4,5,6,7,8,9. Here we report an organoid culture system that generates complex skin from human pluripotent stem cells. We use stepwise modulation of the transforming growth factor β (TGFβ) and fibroblast growth factor (FGF) signalling pathways to co-induce cranial epithelial cells and neural crest cells within a spherical cell aggregate. During an incubation period of 4–5 months, we observe the emergence of a cyst-like skin organoid composed of stratified epidermis, fat-rich dermis and pigmented hair follicles that are equipped with sebaceous glands. A network of sensory neurons and Schwann cells form nerve-like bundles that target Merkel cells in organoid hair follicles, mimicking the neural circuitry associated with human touch. Single-cell RNA sequencing and direct comparison to fetal specimens suggest that the skin organoids are equivalent to the facial skin of human fetuses in the second trimester of development. Moreover, we show that skin organoids form planar hair-bearing skin when grafted onto nude mice. Together, our results demonstrate that nearly complete skin can self-assemble in vitro and be used to reconstitute skin in vivo. We anticipate that our skin organoids will provide a foundation for future studies of human skin development, disease modelling and reconstructive surgery.
Subscribe to Journal
Get full journal access for 1 year
only $3.90 per issue
All prices are NET prices.
VAT will be added later in the checkout.
Rent or Buy article
Get time limited or full article access on ReadCube.
All prices are NET prices.
The scRNA-seq data have been uploaded to the Gene Expression Omnibus with accession code GSE147206. In the Supplementary Data, we have provided HTML files that generate an online interface with which to explore our scRNA-seq analysis pipeline and evaluate additional cell cluster markers for the integrated and non-integrated datasets (Supplementary Data 1–7). In addition, the integrated datasets shown in Fig. 2 can be queried via a data exploration portal hosted on the Koehler laboratory website: https://www.koehler-lab.org/resources. Source data for Figs. 1–4 are provided with the Article.
Scripts used for single-cell RNA-sequencing analysis are available at https://github.com/Koehler-Lab/Lee_et_al_2020_Nature.
Sun, B. K., Siprashvili, Z. & Khavari, P. A. Advances in skin grafting and treatment of cutaneous wounds. Science 346, 941–945 (2014).
Lee, J. et al. Hair follicle development in mouse pluripotent stem cell-derived skin organoids. Cell Rep. 22, 242–254 (2018).
Yang, R. et al. Generation of folliculogenic human epithelial stem cells from induced pluripotent stem cells. Nat. Commun. 5, 3071 (2014).
Abaci, H. E. et al. Tissue engineering of human hair follicles using a biomimetic developmental approach. Nat. Commun. 9, 5301 (2018).
Gledhill, K. et al. Melanin transfer in human 3D skin equivalents generated exclusively from induced pluripotent stem cells. PLoS ONE 10, e0136713 (2015).
Itoh, M. et al. Generation of 3D skin equivalents fully reconstituted from human induced pluripotent stem cells (iPSCs). PLoS ONE 8, e77673 (2013).
Lei, M. et al. Self-organization process in newborn skin organoid formation inspires strategy to restore hair regeneration of adult cells. Proc. Natl Acad. Sci. USA 114, E7101–E7110 (2017).
Heitman, N., Saxena, N. & Rendl, M. Advancing insights into stem cell niche complexities with next-generation technologies. Curr. Opin. Cell Biol. 55, 87–95 (2018).
Koehler, K. R. et al. Generation of inner ear organoids containing functional hair cells from human pluripotent stem cells. Nat. Biotechnol. 35, 583–589 (2017).
Koehler, K. R., Mikosz, A. M., Molosh, A. I., Patel, D. & Hashino, E. Generation of inner ear sensory epithelia from pluripotent stem cells in 3D culture. Nature 500, 217–221 (2013).
Tchieu, J. et al. A modular platform for differentiation of human PSCs into all major ectodermal lineages. Cell Stem Cell 21, 399–410 (2017).
Wilson, P. A. & Hemmati-Brivanlou, A. Induction of epidermis and inhibition of neural fate by Bmp-4. Nature 376, 331–333 (1995).
Minoux, M. & Rijli, F. M. Molecular mechanisms of cranial neural crest cell migration and patterning in craniofacial development. Development 137, 2605–2621 (2010).
Driskell, R. R. et al. Distinct fibroblast lineages determine dermal architecture in skin development and repair. Nature 504, 277–281 (2013).
Betters, E., Liu, Y., Kjaeldgaard, A., Sundström, E. & García-Castro, M. I. Analysis of early human neural crest development. Dev. Biol. 344, 578–592 (2010).
Lee, R. T. et al. Cell delamination in the mesencephalic neural fold and its implication for the origin of ectomesenchyme. Development 140, 4890–4902 (2013).
Driskell, R. R., Giangreco, A., Jensen, K. B., Mulder, K. W. & Watt, F. M. Sox2-positive dermal papilla cells specify hair follicle type in mammalian epidermis. Development 136, 2815–2823 (2009).
Chuong, C.-M., Yeh, C.-Y., Jiang, T. X. & Widelitz, R. Module-based complexity formation: periodic patterning in feathers and hairs. Wiley Interdiscip. Rev. Dev. Biol. 2, 97–112 (2013).
Velasco, S. et al. Individual brain organoids reproducibly form cell diversity of the human cerebral cortex. Nature 570, 523–527 (2019).
Pijuan-Sala, B. et al. A single-cell molecular map of mouse gastrulation and early organogenesis. Nature 566, 490–495 (2019).
Soldatov, R. et al. Spatiotemporal structure of cell fate decisions in murine neural crest. Science 364, eaas9536 (2019).
Joost, S. et al. Single-cell transcriptomics reveals that differentiation and spatial signatures shape epidermal and hair follicle heterogeneity. Cell Syst. 3, 221–237 (2016).
Zeisel, A. et al. Molecular architecture of the mouse nervous system. Cell 174, 999–1014 (2018).
Sennett, R. et al. An integrated transcriptome atlas of embryonic hair follicle progenitors, their niche, and the developing skin. Dev. Cell 34, 577–591 (2015).
Lanctôt, C., Moreau, A., Chamberland, M., Tremblay, M. L. & Drouin, J. Hindlimb patterning and mandible development require the Ptx1 gene. Development 126, 1805–1810 (1999).
Minoux, M. et al. Mouse Hoxa2 mutations provide a model for microtia and auricle duplication. Development 140, 4386–4397 (2013).
Lim, X. & Nusse, R. Wnt signaling in skin development, homeostasis, and disease. Cold Spring Harb. Perspect. Biol. 5, a008029 (2013).
Zhu, X.-J. et al. BMP-FGF signaling axis mediates Wnt-induced epidermal stratification in developing mammalian skin. PLoS Genet. 10, e1004687 (2014).
Richardson, G. D. et al. KGF and EGF signalling block hair follicle induction and promote interfollicular epidermal fate in developing mouse skin. Development 136, 2153–2164 (2009).
Langbein, L., Yoshida, H., Praetzel-Wunder, S., Parry, D. A. & Schweizer, J. The keratins of the human beard hair medulla: the riddle in the middle. J. Invest. Dermatol. 130, 55–73 (2010).
Gatto, G., Smith, K. M., Ross, S. E. & Goulding, M. Neuronal diversity in the somatosensory system: bridging the gap between cell type and function. Curr. Opin. Neurobiol. 56, 167–174 (2019).
Perdigoto, C. N., Bardot, E. S., Valdes, V. J., Santoriello, F. J. & Ezhkova, E. Embryonic maturation of epidermal Merkel cells is controlled by a redundant transcription factor network. Development 141, 4690–4696 (2014).
Jenkins, B. A. & Lumpkin, E. A. Developing a sense of touch. Development 144, 4078–4090 (2017).
Narisawa, Y., Hashimoto, K., Nakamura, Y. & Kohda, H. A high concentration of Merkel cells in the bulge prior to the attachment of the arrector pili muscle and the formation of the perifollicular nerve plexus in human fetal skin. Arch. Dermatol. Res. 285, 261–268 (1993).
Toyoshima, K. E. et al. Fully functional hair follicle regeneration through the rearrangement of stem cells and their niches. Nat. Commun. 3, 784 (2012).
Horsley, V., Aliprantis, A. O., Polak, L., Glimcher, L. H. & Fuchs, E. NFATc1 balances quiescence and proliferation of skin stem cells. Cell 132, 299–310 (2008).
Roberts, B. et al. Systematic gene tagging using CRISPR/Cas9 in human stem cells to illuminate cell organization. Mol. Biol. Cell 28, 2854–2874 (2017).
Lee, J. & Koehler, K. R. Generation of human hair-bearing skin organoids from stem cells. Protoc. Exch. https://doi.org/10.21203/rs.3.pex-889/v1 (2020).
Butler, A., Hoffman, P., Smibert, P., Papalexi, E. & Satija, R. Integrating single-cell transcriptomic data across different conditions, technologies, and species. Nat. Biotechnol. 36, 411–420 (2018).
Satija, R., Farrell, J. A., Gennert, D., Schier, A. F. & Regev, A. Spatial reconstruction of single-cell gene expression data. Nat. Biotechnol. 33, 495–502 (2015).
Stuart, T. et al. Comprehensive integration of single-cell data. Cell 177, 1888–1902 (2019).
Hama, H. et al. ScaleS: an optical clearing palette for biological imaging. Nat. Neurosci. 18, 1518–1529 (2015).
This work was supported by the Ralph W. and Grace M. Showalter Trust (K.R.K.), the Indiana Clinical and Translational Sciences Institute (core pilot grant UL1 TR001108 to K.R.K. and A. Shekhar), the Indiana Center for Biomedical Innovation (Technology Enhancement Grant to K.R.K.) and the NIH (grants R01AR075018, R01DC017461 and R03DC015624 to K.R.K.). Cell lines associated with this study were stored in a facility constructed with support from the NIH (grant C06 RR020128-01). The University of Washington Birth Defects Research Laboratory was supported by NIH award number 5R24HD000836 from the Eunice Kennedy Shriver National Institute of Child Health and Human Development. We thank B. Koh, W. van der Valk, D. O’Day, I. Glass, A. Tward, S. Frumm, D. Spandau, J. Foley, U. Arimpur, E. Longworth-Mills, P.-C. Tang, A. Elghouche, M. Kamocka and C. Miller for their technical assistance, and M. Rendl and N. Saxena for critical comments on the manuscript.
J.L. and K.R.K, with the Indiana University Research and Technology Corporation, are inventors on a published patent covering the entire skin organoid induction method and including much of the data presented in this Article (WO2017070506A1). The other authors declare no competing interests.
Peer review information Nature thanks Claire Higgins, Valerie Horsley and the other, anonymous, reviewer(s) for their contribution to the peer review of this work.
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Extended data figures and tables
Extended Data Fig. 1 Overview of skin organoid induction in relation to developmental events in normal human skin and morphological changes in skin organoids under different treatment regimes.
a, Schematic overview of the skin organoid protocol. (i) Pre-aggregate human pluripotent stem cells to form an aggregate; (ii) induce differentiation of the aggregate to form surface ectoderm and CNC cells by modulating the TGFβ, BMP, and FGF signalling pathways; and (iii) provide an appropriate environment for the aggregate to mature into a complex skin organoid unit composed of fully stratified skin with a dermal layer that produces hair follicles, sensory neurons and cartilage. The cystic skin organoid can be transplanted and integrated into the back skin of a mouse as a planar layer, and can grow hair follicles. b, Comparison of the timeline of skin development in vitro and in vivo. Developmental timing is approximate. SkO, skin organoid. c, Representative differential interference contrast (DIC) and endogenous GFP fluorescence images of WA25 and DSP-GFP skin organoids on different days of differentiation under different treatment regimes. Top, WA25 skin organoids were differentiated in E6-based medium, treated with one of three different regimes on day 3—no LDN or FGF (d0BSF); LDN alone (d0BSF-d3L); and both LDN and FGF (d0BSF-d3LF)—and monitored during days 0–60. The aggregates that were treated with neither LDN nor FGF on day 3 maintained cystic organoids for about 30 days of differentiation, but the organoids lost their morphology and shrunk afterwards. Treatment with LDN alone on day 3 promoted cystic morphology, but was not sufficient to produce hair-bearing skin organoids in a consistent manner. Co-treatment with LDN and FGF on day 3 (the final optimized treatment condition) was optimal for epithelial stratification and sufficient development of the dermal layer. Tail-like regions containing mesenchymal and neuronal cells on one pole of the skin organoids are visible by day 18 of differentiation. Bottom, DSP-GFP skin organoids were also differentiated in E6-based medium and were treated on day 3 with both LDN and FGF (final optimized treatment regime). GFP+ epithelium is visible on the surface or edge of the sphere-like organoids, and the GFP signal intensifies as the organoids differentiate and mature further. The tail portion of the organoid appears by day 22 of differentiation (the GFP+ signal at the tail portion presented in the day-22 image is an autofluorescence). Right, WA25 skin organoids were differentiated in chemically defined medium (CDM)-based medium and treated with one of three different regimes on day 4—no LDN or FGF (NT); LDN alone (d4 LDN); and both LDN and FGF (d4LDN/FGF). By day 55 of differentiation, aggregates that were treated with neither LDN nor FGF on day 4 eventually lost their shape and shrunk. Treatment with LDN alone on day 4 maintained cystic organoid morphology, and occasionally induced maturation of skin organoids. Co-treatment with LDN and FGF on day 4 was suitable for skin organoids to fully mature in a relatively consistent manner. Changing the differentiation medium from CDM to E6 improved skin organoid development by reducing the tail portion (non-skin-related mesenchymal cells) and increasing the head (skin) portion. Optimization experiments were repeated at least three times independently. Scale bars, 500 μm. This figure corresponds with the data and concepts in Fig. 1.
Extended Data Fig. 2 Induction of surface ectoderm and CNC cells in WA25 and DSP-GFP skin organoid cultures.
a–d, Representative immunostaining images of day-16 WA25 (a, b) and DSP-GFP (c, d) aggregates with co-induced epithelium and CNC cells cultured in the optimized d0BSF-d3LF treatment conditions. P75 and SOX10 highlight neuroglia-associated CNC cells, and PDGFRα and TFAP2A highlight mesenchyme-associated CNC cells. In a, b, the epithelium is highlighted by ECAD and TFAP2A double-positive signals. In c, d, ECAD was omitted to allow observation of the endogenous DSP-GFP signal at the apical surface of the epithelium. Dashed boxes indicate magnified regions (right). Dashed lines outline the borders between the epithelium and CNC cells and between neuroglia- and mesenchyme-associated CNC cells. Note that there are non-specific fluorescence background noise signals from Matrigel. The immunostaining was repeated three times independently on a total of nine aggregates from three separate experiments. Scale bars: 100 μm (main images), 50 μm (magnified images). This figure corresponds with the data and concepts in Fig. 1c.
Extended Data Fig. 3 Comparison of initial hair follicle induction in WA25 and DSP-GFP live-cell aggregates.
a, Representative DIC images of WA25 skin organoids at days 65–130 with developing hair follicles. Scale bars, 500 μm (left); 250 μm (right). b, DIC and endogenous GFP fluorescence images of DSP-GFP skin organoids at days 65–135 with developing hair follicles. Dashed boxes indicate magnified regions (right). Scale bars, 500 μm (left, middle); 250 μm (far right, days 95, 110 and 135); 100 μm (far right, days 65 and 85). c, DIC images of WA01 skin organoids at days 68–98 with developing hair follicles. Right, magnified view of day-88 WA01 skin organoid hair follicles. The head (skin cyst) and tail (non-skin mesenchymal) portions within skin organoids are distinguishable. In addition, the hair bulbs are facing outward from the organoid, contacting the medium, whereas the hair shafts grow inward towards the centre of the skin organoid. Scale bars, 250 μm (all panels except far right); 50 μm (far right). The WA25 and DSP-GFP skin organoid images represent morphologies observed throughout nine independent cultures, and the WA01 skin organoid images represent one experiment performed by the Stanford group. This figure corresponds with the data in Fig. 1e, f.
a, A day-75 WA25 skin organoid with nascent hair follicles immunostained for KRT5, KRT17 and TFAP2A. This KRT17 antibody labels basal and peridermal keratinocytes. KRT5 and TFAP2A double-positive staining represents the basal layer, and TFAP2A single-positive staining between basal and peridermal layers indicates the intermediate layer. Dashed boxes indicate magnified regions. Scale bars, 100 μm. b, Immunostaining for ECAD in day-75 WA25 skin organoids labels the entire epithelium, whereas PCAD expression is restricted to the basal layer and the hair germ epithelium. LHX2 labels hair placodes and matrix cells. Dashed boxes indicate magnified regions of the hair germ images (top right insets). Scale bars, 250 μm (far left); 100 μm (right three panels); 23 μm (insets). c, Immunostaining for EDAR in day-75 WA25 skin organoids labels placodes and matrix cells of hair follicles. Scale bars, 50 μm. d, Low-magnification immunostaining image of a day-75 WA25 organoid shows the distribution of KRT5 and TFAP2A expression, representing basal and intermediate layers. The non-specific fluorescence signal in the core of the skin cyst is cell debris. Asterisks indicate developing hair follicle placodes and germs. Scale bar, 100 μm. e, Representative immunostaining image for PDGFRα and SOX2 in a day-85 WA25 skin organoid. PDGFRα is expressed throughout the dermis that has developed in the outer layer of the skin cyst, and SOX2 labels dermal condensate and dermal papilla cells, yet is also expressed in some basal epidermal cells—probably Merkel cells. Scale bar, 50 μm. f, By day 75 of differentiation, NPNT is localized to the basement membrane of the epithelium of WA25 skin organoids. Scale bars, 100 μm. g, Immunostaining image of a day-147 WA25 skin organoid. Expression of αSMA (alpha smooth muscle actin) and ITGA8 (integrin alpha 8)—characteristics of arrector pili muscle—was detected next to the NPNT+ hair follicle bulge region, suggesting that our skin organoid may be capable of producing arrector pili muscle. However, this feature was observed extremely rarely in our present culture conditions, which are still at an early developmental stage, and further optimization of the medium will be necessary for long-term culture. Dashed boxes indicate magnified regions of the bulge with an arrector pili muscle-like phenotype (right). Scale bars, 100 μm. h, Immunostaining of a day-95 DSP-GFP skin organoid showing a hair peg with KRT17+ peridermal layer and outer root sheath, SOX2+ dermal papilla and GFP+ desmosome-rich intermediate epidermis of the hair follicle. Scale bars, 50 μm. All immunostaining was repeated 3–5 times on 9–12 skin organoids generated from 4 independent experiments before selection of representative images. This figure corresponds with the data in Fig. 1.
Extended Data Fig. 5 scRNA-seq analysis of day-6 and day-29 skin organoids derived from WA25 and DSP-GFP cells.
a, g, Separate UMAP plots for WA25 and DSP-GFP cell clusters at day 6 (a) and day 29 (g). The major cell cluster groupings of surface epithelia, epidermis and mesenchyme are noted. Colours indicate cell state. The presumptive cell identities (based on a priori knowledge of marker genes) are listed. Cell clusters with no discernible identity had low expression of mitochondrial genes (‘low mito cells’) or high expression of long noncoding RNA genes (‘high lncRNA’). Day 6 (a): n = 11,785 WA25 cells, n = 11,544 DSP-GFP cells; day 29 (g): n = 9,268 WA25 cells, n = 9,013 DSP-GFP cells. Ten day-6 organoids from one experiment and five day-29 organoids from one experiment were randomly pooled for scRNA-seq analysis (per cell line). b, Dot plot for surface ectoderm and non-neural ectoderm (NNE) markers on the basis of RNA-seq data from a previous study11. c, UMAP plot for HAND1, a key marker for NNE cells derived from human pluripotent stem cells. d, Dot plot for markers of surface ectoderm and anterior–posterior placodes. Gene expression frequency is indicated by spot size and expression level is indicated by colour intensity. e, UMAP overlay plot showing the distribution of OTX1 (marker of anterior placode and neuroectoderm) and GBX2 (marker of posterior placode). f, UMAP plots for key markers of epidermal progenitors, neuroectoderm, general placode and cycling cells. h, UMAP plots for specific marker genes that define epidermal, cycling, CNC and dermal cell subtypes. i, Magnified view of clusters 9 (CNC cell and Schwann cell precursors), 15 (neuroectoderm cells), 18 (PNS- or CNS-like neurons), 19 (myocytes) and 22 (melanocytes). Key marker genes are shown to label each cell cluster. SOX2 has broad expression across neuroectoderm and CNC cells. j, Dot plot for general mesenchymal markers (PRRX1 and PDGFRA), markers of all pharyngeal arches, markers of PA1 and markers of PA2–PA4. Gene expression frequency is indicated by spot size and expression level is indicated by colour intensity. The expression of HOX genes appears to be limited to a subset of cells with low mitochondrial gene expression. This figure corresponds with the data in Fig. 2.
a, UMAP clustering of day-48 WA25 cell subtypes. Colours indicate cell state. n = 2,491 cells. Six day-48 skin organoids from one experiment were pooled for scRNA-seq analysis. SCPs, Schwann cell precursor cells. b, Heat map showing the scaled log-normalized expression of the top 10 differentially expressed genes per cell cluster for the day-48 WA25 scRNA-seq dataset. ELN is also known as SMIM6. c, Dot plot array showing the top 10 positively expressed genes per cell cluster for the day-48 WA25 scRNA-seq dataset. Gene expression frequency is indicated by spot size and expression level is indicated by colour intensity. d, UMAP plots for marker genes of specific cell subtypes. e, UMAP plots for WNT signalling pathway genes. WNT6 is expressed in basal keratinocytes and peridermal keratinocytes. LEF1 expression appears localized to basal keratinocytes. Negative WNT modulatory genes, SFRP2 and WIF1, are expressed in putative dermal fibroblasts of the mesenchymal cell group. f, Dermal fibroblast clusters also contain cells that express FGF7 (keratinocyte growth factor). The numbers of cell clusters with positive expression are listed on the UMAP plot. g, Identification of Merkel cells. Using the UMAP clustering algorithm, we identified a subset (n = 8 cells) of cells of cluster 0 (putative basal keratinocytes) that were completely separated from most cluster-0 cells, suggesting that our unbiased analysis pipeline failed to identify a unique subset of low-abundance cells. We used the Seurat manual selection tool to generate an 11th cluster containing these cells (a, left). Violin plots show normalized gene expression of the Merkel cell marker genes ATOH1, ISL1, SOX2, KRT8, KRT18 and KRT20.
a, TEM image of epidermis in a day-140 WA25 skin organoid. A melanocyte cell body is pseudo-coloured pink, and the basal skin layer is pseudo-coloured green. BM, basement membrane. Scale bar, 10 μm. b, Higher-magnification image of basal keratinocytes and the basement membrane (top) and spinous layer keratinocytes (bottom). Scale bars, 5 μm. c, Higher-magnification image of matrix-associated melanocytes (pink) with melanosomes at different stages (I, II, III and IV). The region in the dashed box is magnified in the lower panel. Scale bars, 5 μm (top); 800 nm (bottom). The images in c correspond to Fig. 3f. TEM was performed on two day-140 skin organoids from separate experiments (a–c). d, Dark-field illumination image of day-124 WA25 skin organoids, comparing the development of pigmented and non-pigmented (albino) hairs. Images represent pigmented and non-pigmented hairy skin organoids from nine independent experiments. Scale bar, 1 mm. e, Representative dark-field illumination (left, one asterisk) and endogenous GFP fluorescence (right, two asterisks) images of day-125 DSP-GFP skin organoids with pigmented hair follicles (dashed boxes in f, g). Scale bars, 1 mm. f, g, Overview of representative 16 out of 24 skin organoids from one DSP-GFP experiment. Each organoid exhibits pigmented hair follicles; however, the morphology of the epidermal cyst, as shown by DSP-GFP expression (green), varies between organoids. Magnified views of the organoids highlighted with dashed boxes and asterisks are shown in e. Scale bars, 1 mm. Images show a set of organoids from one experiment and are representative of morphologies observed over the course of nine independent experiments (e–g). h, Immunostaining for TUJ1 and ISL1 in a day-140 WA25 organoid reveals the ganglion-like cluster of neurons. Scale bar, 50 μm. i, A subset of organoid neurons express NEFH and are associated with S100β+ satellite glial cells and Schwann cells. Dashed circles indicate axons of neurons. Scale bars, 50 μm (main image); 25 μm (inset). j, Immunostaining for PVALB in a day-125 WA25 organoid reveals the proprioceptors. PRPH+ sensory neurons have small somas, and their axons form a fascicle. Scale bar, 35 μm. Immunostaining was repeated at least three times on five or six independent organoids (h–j). This figure corresponds with the data in Fig. 3.
a, b, Violin and t-SNE plots showing normalized expression of chondral marker genes within cluster 7 of the day-48 WA25 skin organoid dataset (Extended Data Fig. 6, Supplementary Data 7). LECT1 is also known as CNMD. The data represent cells pooled from six day-48 skin organoids from one experiment. c, Haematoxylin-stained section of a day-140 skin organoid showing hyaline cartilage that has formed within skin organoid-associated mesenchymal tissue. Three independent haematoxylin stainings were performed on six skin organoids from three different experiments. d, TEM image of two representative chondrocytes located within hyaline cartilage tissue (dashed box in c). TEM was performed once on two different skin organoids. e, f, Immunostaining of day-140 organoid samples for aggrecan (ACAN) and collagen type II alpha 1 (COL2A1) highlights cartilage development. Images represent one of three independent IHC stainings on six skin organoids produced from three separate experiments. Scale bars, 100 μm (c–f). This figure corresponds with the data in Figs. 2, 3.
Extended Data Fig. 9 Hair follicles of xenografted WA25 skin organoids have sebaceous glands comparable to second-trimester (18 weeks) fetal and adult facial hair.
a, Bright-field image of two day-115 (in vitro) WA25 skin organoid hair follicles with visible hair shafts and sebaceous glands. Image represents hair follicles produced in nine independent experiments. Scale bar, 250 μm. b, Left, LipidTOX staining of day-120 WA25 skin organoid hair follicles reveals lipid-rich cells such as sebocytes and adipocytes (outlined by dashed lines). Immunostaining for KRT15 labels the outer root sheath of hair follicle cells. Right, high-magnification image of an adjacent cryosection of the specimen in the left panel. Dashed line highlights a horizontally cross-sectioned hair shaft and a sebaceous gland. Scale bars, 100 μm (left); 50 μm (right). c, TEM image of a sebaceous gland in a day-140 WA25 skin organoid. TEM was performed once on two different skin organoids from separate experiments. Peripheral layer cells (PLCs) and a maturing sebocyte (MS) containing sebum vacuoles (SV) have been pseudo-coloured in yellow and green, respectively. Scale bar, 5 μm. d, SCD1+ sebaceous glands in xenografted day-140 WA25 skin organoids. Xenografts were extracted and fixed in 4% PFA, more than 49 days after transplantation. Dashed boxes and asterisks indicate the magnified regions in the subsequent images. Dashed lines distinguish between the tissue of the grafted human skin organoid and the tissues of the host NU/J mouse. Arrows indicate SCD1+ adipocytes. Scale bars, 100 μm (d, d’); 50 μm (*, **, separate cryosection). e, SCD1+ sebaceous glands in the NU/J mouse skin adjacent to xenografts. The size of mouse sebaceous glands is smaller than that of human sebaceous glands; the origin of sebaceous glands is distinguishable within the extracted xenograft samples. Scale bar, 50 μm. f, SCD1+ sebaceous glands in 18-week human fetal forehead skin. Scale bars, 250 μm (f); 100 μm (f’); 50 μm (*, separate cryosection). g, SCD1+ sebaceous glands and adipocytes in 18-week human fetal cheek skin. The adipocytes are prominently abundant in fetal cheek tissue compared to forehead tissue (f versus g). Dashed boxes and symbols highlight the magnified regions in the subsequent images. Scale bars, 250 μm (g); 100 μm (g’, **); 50 μm (*, #, ##, separate cryosection). h, i, SCD1+ sebaceous glands in adult human facial skin. The epithelium is visualized by ECAD immunostaining. Dashed lines outline several lobes of the sebaceous glands of a hair follicle. Scale bars, 250 μm (h); 100 μm (h’, i). Immunostaining images represent one of five independent stainings on three to five different samples per tissue type. This figure corresponds with the data in Fig. 4.
Extended Data Fig. 10 Hair follicles of xenografted WA25 skin organoids have bulge regions comparable to second-trimester (18 weeks) fetal and adult facial hair.
a–d, Representative immunostaining images for markers of hair follicle stem cells (LHX2, KRT15 and NFATC1) in the hair follicle bulge region in 18-week human fetal skin from two facial locations (forehead and cheek) (a, b), adult facial skin (c) and xenografted skin organoid tissue (d). In both fetal and xenograft hair follicles, NFATC1 expression is predominantly localized to the cytoplasm in bulge cells, whereas in adult facial hair follicles, NFATC1 expression is nuclear-localized in the hair follicle bulge region—consistent with previous reports of nuclear-localized NFATC1 in mouse bulge stem cells36. Arrows indicate background (BG) staining noises. Dashed boxes in a, c, d indicate magnified bulge regions, which are presented in a corner of each image. Dashed brackets in b indicate the bulge region. Representative immunostaining images are selected from five independent stainings on three to five different samples per tissue type. Scale bars, 100 μm (a, c); 50 μm (b, d). This figure corresponds with the data in Fig. 4.
This file contains a Supplementary Note, Supplementary Discussion and Supplementary References.
Supplementary Table 1 | Comparison of hair follicle formation frequencies between cell lines and treatments.
Supplementary Table 2 | Media compositions.
Supplementary Table 3 | Antibodies list.
This file contains Supplementary Data files 1-7 and a Supplementary Data Guide.
Wholemount skin organoid immunostained with antibodies for KRT5 (red) and SOX2 (cyan). The sample was cleared using the ScaleS approach and imaged on an Olympus FV1000 confocal microscope. The dense clusters of SOX2+ cells represent the dermal condensate and dermal papilla cells. The SOX2+ cells embedded in the outer root sheath of each hair follicle are presumed to be Merkel cells, while the extra-epithelial SOX2+ cells are presumed to be melanoblasts.
Wholemount skin organoid immunostained with antibodies for KRT5 (red), SOX2 (cyan), and TUJ1 (green). Image segmentation using the Imaris “spots” module to estimate that the dermal papilla (DP) cell clusters contained approximately 250, 335, and 522 cells. Note that the TUJ1+ neurites are innervating/wrapping around hair follicle bulbs where dermal papilla cells are located. Also, note that SOX2+ Merkel cells are TUJ1+.
Wholemount skin organoid immunostained with an antibody for PMEL (red). Premelanosome protein (PMEL). PMEL+ melanocytes are scattered throughout the epidermis and hair follicle outer root sheath and concentrated at the hair follicle bulbs where matrix cells are located.
Wholemount skin organoid immunostained with an antibody for TUJ1 (red). Hoechst staining for nuclei is displayed as cyan. Closeup view of a neural ganglion where TUJ1+ soma containing satellite glial cells and Schwann cells with elongated nuclei are clearly visible. Note that the neurons are morphologically pseudo-unipolar neurons. Also, note that the fasciculate TUJ1+ neurons are stretched out throughout the organoid, interweaving between hair follicles.
Wholemount skin organoid immunostained with antibodies for NEFH (green) and PRPH (red). The skin organoid ganglion has neuronal diversity. Peripherin (PRPH) labels sensory neurons; small soma neurons are PRPH+NEFH-, while large soma neurons are PRPH+NEFH+.
Wholemount skin organoid immunostained with antibodies for KRT17 (red) and TUJ1 (green). Skin organoid hemispheres in the first part of this video provide an overview of neurons in an organoid and highlight representative innervation sites. Higher magnification views of an innervation site shows neurite targeting of a hair follicle upper bulge (Isthmus) region where Merkel cells reside. Depth coding reveals that the Merkel cells are located in the outer root sheath of hair follicle, underneath the layer of neurites.
Wholemount skin organoid immunostained with antibodies for KRT17 (red) and TUJ1 (green). Another representative video showing the innervation sites of TUJ1+ neurons. The nerve endings of neurons clearly target the hair follicle bulge region.
Wholemount skin organoid immunostained with antibodies for KRT17 (red), KRT20 (green), and NPNT (white). NPNT is expressed in the basal membrane of epidermis. Note the high cell density of outer root sheath cells in the bulge region compared to non-bulge region (upper part of the frame). KRT20 labels Merkel cells.
About this article
Cite this article
Lee, J., Rabbani, C.C., Gao, H. et al. Hair-bearing human skin generated entirely from pluripotent stem cells. Nature 582, 399–404 (2020). https://doi.org/10.1038/s41586-020-2352-3