Restoration of keratinocytic phenotypes in autonomous trisomy-rescued cells

An extra copy of chromosome 21 in humans can alter cellular phenotypes as well as immune and metabolic systems. Down syndrome is associated with many health-related problems and age-related disorders including dermatological abnormalities. However, few studies have focused on the impact of trisomy 21 (T21) on epidermal stem cells and progenitor cell dysfunction. Here, we investigated the differences in keratinocytic characteristics between Down syndrome and euploid cells by differentiating cells from trisomy 21-induced pluripotent stem cells (T21-iPSCs) and autonomous rescued disomy 21-iPSCs (D21-iPSCs). Our protocol for keratinocytic differentiation of T21-iPSCs and D21-iPSCs was employed. For propagation of T21- and D21-iPSC-derived keratinocytes and cell sheet formation, the culture medium supplemented with Rho kinase inhibitor on mouse feeder cells was introduced as growth rate decreased. Before passaging, selection of a keratinocytic population with differential dispase reactivity was performed. Three-dimensional (3D) air-liquid interface was performed in order to evaluate the ability of iPSC-derived keratinocytes to differentiate and form stratified squamous epithelium. Trisomy-rescued disomy 21-iPSCs were capable of epidermal differentiation and expressed keratinocytic markers such as KRT14 and TP63 upon differentiation compared to trisomy 21-iPSCs. The lifespan of iPSC-derived keratinocytes could successfully be extended on mouse feeder cells in media containing Rho kinase inhibitor, to more than 34 population doublings over a period of 160 days. Dispase-based purification of disomy iPSC-derived keratinocytes contributed epidermal sheet formation. The trisomy-rescued disomy 21-iPSC-derived keratinocytes with an expanded lifespan generated 3D skin in combination with a dermal fibroblast component. Keratinocytes derived from autonomous trisomy-rescued iPSC have the ability of stratification for manufacturing 3D skin with restoration of keratinocytic functions.


Background
Down syndrome patients have an extra copy of chromosome 21, which causes health-related problems and agerelated disorders. Trisomy 21 results in the triplication of over 400 genes, which makes clarification of the precise mechanisms of some phenotypes difficult in Down syndrome [1]. The complications of Down syndrome are numerous; clinical manifestations include mental retardation, congenital heart disease, impaired cognition, premature aging, metabolic problems, and leukemia [2,3]. Trisomy 21 is also associated with an increased incidence of several dermatological conditions. Immunological disturbances and/or barrier dysfunction of skin are associated with psoriasis, atopic dermatitis, and cutaneous infections. Keratodermatoses and abnormality of elastic fibers also contribute to dermatological disorders such as anetoderma and keratosis pilaris in Down syndrome patients [4][5][6]. Genetic signatures of these dermatological diseases have been reported using keratinocytes derived from patient-specific iPSCs [7].
A Down syndrome mouse model exhibits hyperproliferation of trisomy 21-keratinocytes in the skin [8]. Cell culture models of trisomy 21 are important for development of conventional therapy for Down syndrome. However, few studies have focused on differences in epidermal characteristics between Down syndrome and euploid cells in vitro. Generation of disease-specific induced pluripotent stem cells (iPSC) allows an alternative paradigm for modeling human genetic disease in vitro [9,10]. Directed differentiation from these iPSC theoretically enables researchers to derive any given diseaseaffected human cell-type [11]. Disease-specific iPSCs, therefore, serve as a powerful tool for investigating the mechanisms of Down syndrome [9,12,13].
The differences in genetic backgrounds of individuals has hindered the assessment the phenotypic disparities between Down syndrome cells and healthy control cells. Paired disomic/trisomic iPSCs derived from the same trisomic individual by insertion of a transgene into a chromosome 21 has allowed investigation of trisomy 21dependent gene expression and the impact of an extra copy of chromosome 21 without the confounding effects of a different genetic background [13][14][15][16]. The artificial selection of disomic clones possibly results in epigenetic changes that add to the transcriptional differences between the trisomic and disomic iPSCs [17]. In the previous study, we have obtained revertant cells with normal chromosome 21 diploids from the trisomy 21 iPSCs during long-term cultivation [18]. Autonomously rescued disomic cells serve as a good control for trisomic cells. Both disomy 21 iPSC and trisomy 21 iPSC have been shown to differentiate into neural stem cells [18].
In this study, we successfully extended the lifespan of trisomy-rescued iPSC-derived keratinocytes to more than 34 population doublings over a period of 160 days with maintenance of keratinocytic phenotypes. In addition, keratinocytes derived from autonomous trisomy-rescued iPSCs regained capability of stratification and were suitable for manufacturing three-dimensional skin with keratinocytic function.

Real-Time qPCR
RNA was extracted from cells using the RNeasy Plus Mini kit (Qiagen, #74134). An aliquot of total RNA was reverse transcribed using an oligo (dT) primer (Invitrogen, #18418-020). For the thermal cycle reactions, the cDNA template was amplified (Applied Biosystems Quantstudio 12K Flex Real-Time PCR System) with gene-specific primer sets (see Additional file 1: Table S1) using the Platinum SYBR Green qPCR SuperMix-UDG with ROX (Invitrogen, #11733-046) under the following reaction conditions: 40 cycles of PCR (95°C for 15 s and 60°C for 1 min) after an initial denaturation (95°C for 2 min). Fluorescence was monitored during every PCR cycle at the annealing step. mRNA levels were normalized using glyceraldehyde-3-phosphate dehydrogenase (GAPDH) as a housekeeping gene. RT-qPCR analyses for expression of epithelial markers were showed as means of percent of GAPDH (%GAPDH) with the standard deviation (SD).

Immunocytochemical analysis
Cells were fixed with 4% paraformaldehyde in PBS for 10 min at 4°C. After washing with PBS and treatment with 0.1% Triton X-100 (Sigma-Aldrich, # T8787-100 ML) for 10 min at room temperature, the cells were incubated with Protein Block Serum-Free Ready-to-use (Dako, # X0909) for 30 min at room temperature. This was followed by reaction with primary antibodies in blocking buffer for 60 min at room temperature. After washing with PBS, the cells were incubated with fluorescently conjugated secondary antibodies; anti-rabbit or anti-mouse immunoglobulin G (IgG) bound to Alexa 488 or 546 (1:1,000) was incubated in blocking buffer for 30 min at room temperature. The nuclei were stained with DAPI (Biotium, # 40043). All images were captured using confocal microscopy (LSM 510 and LSM 510 MERA laser scanning microscope; Carl Zeiss, Germany). The fluorescent cells were quantified by hybrid cell counting using a digital microscope (BZ-X 710; Keyence Corp., Osaka, Japan). We measured the number of positive cells in at least six fields to obtain a total of more than 450 cells. Antibody information is provided in Additional file 2: table S2a.

Immunohistochemical analysis
Immunohistochemistry was performed as previously described [21]. Paraffin sections were deparaffinized and heated in antigen retrieval solution (Nichirei, #415211) for 20 min. After washing with PBS, samples were placed in 3% hydrogen peroxide/methanol for 5 min to block endogenous peroxidase. The sections were then incubated at room temperature for 90 min in primary antibodies diluted with antibody diluent. Antibody information is provided in Additional file 2: table S2b. The sections were then washed three times with PBS and incubated with peroxidase-labeled goat anti-mouse or anti-rabbit antibodies (Nichirei, #424151) at room temperature for 45 min. After washing with PBS, they were incubated in 3,3′diaminobenzidin (Muto pure chemicals, #40651) for 5-10 min. Negative controls were performed by omitting the primary antibody. The sections were counterstained with hematoxylin.

Dispase-based keratinocytic selection
Before passaging, selection of a keratinocytic population with differential dispase reactivity was performed. Coculture of iPSC-derived keratinocytes and MEF under culture condition B were rinsed with PBS and incubated with 10 mg/mL dispase at 37°C for 5-8 min, with close monitoring by phase microscopy. When non-keratinocytes began to detach by tapping the dish, dispase was inactivated, and the solution containing non-keratinocytes and feeder cells were removed by aspiration. Stratified colonies remained tightly adherent. The remaining colonies were rinsed with PBS and incubated with trypsin-EDTA solution at 37°C for about 3 min. The cells were passaged to a new plate of irradiated MEF feeder cells at a density of 0.1-0.25 × 10 5 cells/cm 2 .

Generation of a 3D skin equivalent
Three-dimensional skin was generated according to a previously described protocol [25]. Type I collagen (Koken, #IPC-50) and 1 × 10 6 human foreskin fibroblasts (HFFs) were mixed and poured into an untreated 60-mm Petri dish (Falcon, #351007) while cooling and allowed to gel at 37°C for 1 h to prepare the dermal equivalent. The collagen gel was detached from the surface of the Petri dish, and the contraction of the gel was facilitated. The medium was changed every 2 or 3 days for 7 days. iPSC-derived keratinocytes were plated at 0.2 × 10 6 cells inside in a glass ring (IWAKI, #RING-12) on the surface of the contracted collagen gel, which was plated onto polyethylene terephthalate membranes (Corning, #35-3493). iPSC-derived keratinocytes were grown in ESTEM-EP medium for 1 day, following which they were exposed to air in a 1:1 mixture of ESTEM-EP medium and DMEM supplemented with 10% FBS (medium for HFFs), in which the Ca 2+ concentration was adjusted to 0.9 mM. The medium was changed every 2 or 3 days. Multilayered 3D cultures of keratinocytes were obtained by day 14.

Statistical analysis
The results are expressed as means with SD. Comparisons between the two groups were evaluated with the unpaired Student's t test. For three groups, a one-way analysis of variance (ANOVA) with a Bonferroni correction was used. Statistical analysis was performed using Stata 14.0 (StataCorp LP, College Station, Texas, USA). P values less than 0.05 were considered to be statistically significant.
A combination of serum, Y-27632, and feeder cells allows propagation and sheet formation of D21-KCs To escape cell senescence and form epidermal sheets, we altered the cultivation conditions (Fig. 2a). T21-KCs, D21-KC#1, and D21-KC#2 recovered the ability to grow and proliferated as colonies on MEFs (Fig. 2b). The differentiated cells under these culture conditions were defined as mature keratinocytes. We analyzed these cells 1-2 weeks after the change in cultivation conditions. T21-KCs showed multilayer structures with a high nucleus/cytoplasm (N/C) ratio and chromatin aggregation (Fig. 2c). In contrast, D21-KC#1 and D21-KC#2 cells showed an epithelial structure with basal, spinous, and granular layers (Fig. 2d). Uniformly thick multilayersheets were obtained by dispase treatment (Fig. 2e). Immunostaining of D21-KC stratified epithelium revealed layer-dependent reactivity to KRT14, basement membrane integrin β4, the suprabasal marker KRT10, and spinous and granular layer markers involucrin (Fig. 2f). Loricrin, a major component of granular layers, was also detected in the KRT14-negative cells. In contrast to D21-KC#1 and D21-KC#2 cells, T21-KC expressed OCT3/4 in some cells and failed to express KRT10 and KRT14. These results suggest that the autonomously rescued disomy 21 keratinocytes produced epithelial structures. Along with the immunocytochemical analysis, RT-qPCR analysis revealed a lack of expression of epithelial markers KRT14, TP63, and Involucrin in T21-KCs (Fig. 2g). D21-KC#1, but not T21-KCs, significantly increased in the epithelial markers after the change in culture conditions. These significant increases in keratinocytic marker expression are probably due to culture condition B, which is suitable for keratinocyte proliferation and maturation (Fig. 2a). The expression of OCT3/4 and NANOG diminished in both T21-and D21-KCs. Cytological analysis showed that T21-KCs, D21-KC#1, and D21-KC#2 treated with condition A showed variable cell sizes with a low N/C ratios (see Additional file 4: Fig. S2).
Trisomy 21-and rescued disomy 21-iPSCs escape from senescence for more than 160 days The proliferative capability of T21-KC, D21-KC#1, and D21-KC#2 was restored after the change in cultivation conditions (Fig. 3). T21-KC cells continued to proliferate rapidly and had fewer cells that showed keratinocytic morphology. In contrast, D21-KC#1 and D21-KC#2 exhibited stable growth and maintained keratinocytic morphology for a period of more than 160 days. D21-KC#1 and D21-KC#2 showed growth cessation at 34 and 52 PDs, respectively. The reason for the differences in growth cessation is unclear but may be related to the differentiation competency of these cells.

D21-KC#1 keratinocyte-like cells expressed KRT14 in
To separate keratinocytelike cells from non-keratinocyte-like cells, we used differential reactivity of these cells to dispase, i.e., tight or loose cell-dish adhesion: Stratified colonies are relatively resistant to dispase and non-keratinocytes are easily detached after dispase treatment (Fig. 4c). This treatment was applied to D21-KC#1 cells at passage 7 to passage 9 (Fig. 4d). The ratio of keratinocytes increased at each passage by microscopic observation with a phase-contrast microscope. KRT14 increased, TP63 decreased, and Involucrin remained unchanged at each passage ( Fig. 4e-g). D21-KC#1 was immunocytochemically positive for epidermal markers, i.e., KRT14, KRT10, involucrin, and loricrin, at passage 13 (Fig. 4h). D21-KCs were able to reproducibly generate epidermal sheets (Fig. 4i). T21-KCs became fragmented upon dispase treatment; thus, we decided not to use dispase for keratinocytic purification. We picked up colonies with keratinocyte morphology at each passage from passage 6 to passage 8 (see Additional file 5: Fig. S3A, B). Keratinocyte-like cells were microscopically observed at passage 8; however, these cells did not express KRT14 and TP63 and epithelial markers were absent (Fig.  4e, f, h). The lack of keratinocyte marker expression may be caused by a gene dosage imbalance due to trisomy 21.

3D skin can be manufactured from disomy 21 iPSCderived keratinocytes
The dispase-based selection at each passage helped maintain the proliferation of KRT14 positive cells and stratified cell sheets were manufactured in plate cultures of D21-KC#1. We therefore investigated whether disomy 21 iPSC-derived keratinocytes could manufacture 3D skin. To this end, we generated artificial dermis by combining human cultured dermal cells and type I collagen in a Petri dish, and overlaid D21-KC#1 cells in a glass ring. D21-KC#1 cells on the artificial dermis were then cultivated in an air-liquid interface for 2 weeks to accelerate epidermal differentiation. The 3D skin from D21-KC#1 was generated with irregular layer formation and expressed KRT14, KRT10, involucrin, loricrin, and integrin β4 (Fig. 5a, b). Laminin 5 was not detected at the dermal-epidermal junction.

Discussion
Differentiation incompetence in T21-iPSCs and its restoration in D21-iPSCs Restoration of differentiation and growth of trisomyrescued iPSCs is a focus of our laboratory. This restoration or normalization may simply be attributable to a gene dosage effect, i.e., a change from 3 (aneuploid) to 2 copies (euploid). The Down syndrome phenotype is generally derived from the increased expression levels of dosagesensitive genes, and most of the triplicated genes show upregulated expression that is compatible with gene dosage (i.e., at close to 1.5-fold increase) [15,26]. The expression levels of the genes for APP (Alzheimer's disease marker), DYRK1A, DSCR1 (Down syndrome critical region 1), ETS2, and SOD1, all of which are located on chromosome 21, are dysregulated in trisomic cells. Although a gene dosage imbalance is the main molecular mechanism, extensive dysregulation of euploid genes are associated with Down syndrome phenotypes [26,27]. An extra copy of chromosome 21 also alters gene expression across every chromosome, not just chromosome 21 [9,28]. The altered expression levels decrease to normal levels in trisomyrescued undifferentiated iPSCs; in other words, revertant cells regain the gene expression levels of intact iPSCs [18]. These normalized gene expression levels may restore keratinocytic differentiation capability in trisomy-rescued iPSCs. The detailed mechanism of keratinocytic restoration in trisomy 21 cells remains unclear, and additional analysis may elucidate the mechanism underlying epidermal abnormalities in Down syndrome. The ability of trisomy 21-stem/progenitor cells to differentiate also depends on the target organs or cells [14,15,29,30]. The downregulated genes in T21-iPSCs reveal significant enrichment for genes involved in embryonic and tissue morphogenesis [15]. In addition, differentiated ectoderm germ structures are not found in T21-iPSC-derived teratomas [31]. In another study, lack of pluripotency in T21-iPSCs with reduced expression of HERVH might explain the lack of T21-iPSC differentiation compared to disomic iPSCs [15]. These reports may explain the inability of to T21-iPSCs differentiate, i.e., a lack of KRT14 marker and the partial expression of OCT3/4 in T21-KCs, and increased population of nonkeratinocytes after the long-term culture. It is noted that keratinocytic differentiation of D21-iPSCs was achieved at Fig. 3 Differentiating cells from T21-iPSCs and the rescued disomy (D21)-iPSCs shows proliferative capability for more than 160 days. The numbers of cells were recorded at each passage using an automated cell counter and population doublings were determined. The total number of population doublings (PDs) was calculated using the formula [log10 (total number of harvested cells/number of plated cells)]/log10 (2) from two or three independent experiments passage 106. This implies that the differentiation capability of trisomy-rescued cells is extremely stable for a period of more than 500 days.

Uncontrolled growth of T21-KCs and normalized growth of D21-KCs
Accelerated growth of differentiated cells from T21-iPSCs into keratinocytes was unexpected because the incidence of solid tumors, such as dermatological tumors, is decreased in Down syndrome patients [32,33]. Candidate genes such as DYRK1A, PIGP, and RCAN1 are related to keratinocyte hyperproliferation in a Down syndrome mouse model [8]. Triplication of Usp16, located on chromosome 21, reduces self-renewal of hematopoietic stem cells and expansion of mammary epithelial cells, neural progenitors, and fibroblasts [34]. Fig. 4 Purification of iPSC-derived keratinocytes, based on differential reactions to dispase. a Phase-contrast photomicrographs of keratinocytes derived from D21-KC#1 at passage 5. Two types of colonies could be observed: stratified keratinocyte colonies (left) and non-keratinocyte-like cell colonies (right). Scale bars, 200 μm. b Immunocytochemical analysis of D21-KC#1 colony with epithelial markers, i.e., KRT14 and KRT10. Scale bars, 200 μm. c Selection of keratinocytes with dispase digestion. Non-keratinocytes (non-KC) detached earlier than keratinocytes with dispase treatment. Nonkeratinocytes and feeder cells were removed at each passage. Keratinocytes were then passaged. d Phase-contrast photomicrographs of D21-KC#1 during the dispase selection. There were insular colonies, which were formed by stratified epithelial cells, and another area contained non-keratinocytelike cells. Dispase-based selection at each passage caused non-keratinocyte-like cells to decrease. Scale bars, 500 μm. e Real-time qPCR analysis of KRT14 at each passage. Values are shown as means ± SD from two or three independent experiments. f Real-time qPCR analysis of TP63 at each passage. Values are shown as means ± SD from two or three independent experiments. g Real-time qPCR analysis of Involucrin at each passage. Values are shown as means ± SD from two or three independent experiments. h Immunocytochemical analysis of epithelial markers, i.e., KRT14, KRT10, involucrin, and loricrin after colony isolations (T21-KC) at passage 7 and dispase-based selection (D21-KC#1) at passage 13. Scale bars, 100 μm. i Epithelial sheets of keratinocyte-like cells derived from D21-iPSC#1 at passage 8. Cell sheets were harvested by dispase treatment. Scale bar, 1 cm Overexpression of patched-1, a receptor that represses the mitogenic sonic hedgehog pathway, may be associated with proliferation impairment in skin [35][36][37][38]. Moreover, proliferation of trisomy 21-fibroblasts and trisomy 21-iPSCs decrease due to oxidative stress and protein aggregation [39,40]. Neurogenesis-related genes such as DYRK1A and RCAN1, craniofacial defect-related genes such as DYRK1A, ETS2, and RCAN1, and tumor suppressor-related genes such as DYRK1A, ETS2, and RCAN1 were also reported in Down syndrome [26]. In our study, an increased growth rate, an increased nucleus/cytoplasm ratio, and aggregated chromatin of T21-KCs may have resulted from susceptibility to growth stimulation. Skin tumor growth and hyperkeratosis were detected in Ts1Rhr mice model of Down syndrome [8]. We have also reported accelerated growth of T21derived neural stem cells [18]. These increases in growth of other cell types are indeed compatible with that of T21-KCs in this study.
Psoriasis is one of the dermatological abnormalities observed in Down syndrome and is characterized by hyperproliferation and defective differentiation of keratinocytes. Altered gene expression such as upregulation of IFN genes is linked to hyperproliferation of KCs derived from psoriasis patient-specific iPSCs [7]. The autoimmune abnormalities and altered sensitivity to stimuli in Down syndrome may be linked to the excessive proliferation of T21-KCs and reduced differentiation of T21-iPSCs.
In addition to growth rate, senescence or aging of iPSC-derived keratinocytes is a concern. Human primary keratinocytes in serum-free and chemically defined media senesce around 15-20 PDs [41,42]. iPSC-derived keratinocytes exhibit growth arrest when cultured in feeder-free, serum-free medium containing EGF and Y-27632 around passage 4 or 5 [19]. Combination of Rho kinase inhibitor and feeder cells induces conditional reprograming and immortalization of human epithelial cells without the use of viral infection or genetic modification [22][23][24]. In line with these reports, D21-KC survival and proliferation was achieved with a combination of a Rho kinase inhibitor and feeder cells. The successful extension of the D21-KC lifespan to more than 34 population doublings over the period of 160 days is surprising, since extended lifespan of T21-KCs was not accompanied by keratinocytic phenotypes. The trisomy-rescued D21-KCs with a longer lifespan have the ability to stratify, which is essential for manufacturing 3D skin with keratinocytic functions.

Conclusion
To study the impact of trisomy 21 on keratinocytic function, we provide an iPSC-derived model using trisomy 21 iPSC and autonomous trisomy-rescued iPSCs. Our results suggest that there is impairment in keratinocytic differentiation in trisomy 21 iPSCs. In contrast, propagation of trisomy-rescued iPSC-derived keratinocytes with purification and manufacturing of stratified epithelial sheets and 3D skin imply a restoration of keratinocytic functions. Further investigation of the influence of the extra copy of chromosome 21 may help determine the underlying causes of Down syndrome phenotypes and lead to the generation of a Down syndrome model of dermatological abnormalities.