Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2023 May 1.
Published in final edited form as: Dev Biol. 2022 Feb 25;485:9–23. doi: 10.1016/j.ydbio.2022.02.009

Targeted deletion of TGFβ1 in basal keratinocytes causes profound defects in stratified squamous epithelia and aberrant melanocyte migration

Fiona E Chalmers 1, Justyn E Dusold 1, Javed A Shaik 2, Hailey A Walsh 1, Adam B Glick 1,*
PMCID: PMC8969113  NIHMSID: NIHMS1785694  PMID: 35227671

Abstract

Transforming Growth Factor Beta 1 (TGFβ1) is a multifunctional cytokine that regulates proliferation, apoptosis, and epithelial-mesenchymal transition of epithelial cells. While its role in cancer is well studied, less is known about TGFβ1 and regulation of epithelial development. To address this, we deleted TGFβ1 in basal keratinocytes of stratified squamous epithelia. Newborn mice with a homozygous TGFβ1 deletion had significant defects in proliferation and differentiation of the epidermis and oral mucosa, and died shortly after birth. Hair follicles were sparse in TGFβ1 depleted skin and had delayed development. Additionally, the Wnt pathway transcription factor LEF1 was reduced in hair follicle bulbs and nearly absent from the basal epithelial layer. Hemizygous knockout mice survived to adulthood but were runted and had sparse coats. The skin of these mice had irregular hair follicle morphology and aberrant hair cycle progression, as well as abnormally high melanin expression and delayed melanocyte migration. In contrast to newborn TGFβ1 null mice, the epidermis was hyperproliferative, acanthotic and inflamed. Expression of p63, a master regulator of stratified epithelial identity, proliferation and differentiation, was reduced in TGFβ1 null newborn epidermis but expanded in the postnatal acanthotic epidermis of TGFβ1 hemizygous mice. Thus, TGFβ1 is both essential and haploinsufficient with context dependent roles in stratified squamous epithelial development and homeostasis.

Keywords: TGFβ signaling, keratinocyte differentiation, tissue homeostasis, stratified squamous epithelium, melanocytes

Graphical Abstract

graphic file with name nihms-1785694-f0001.jpg

Summary statement

TGFβ1 signaling in basal keratinocytes is essential for maintenance and differentiation of stratified squamous epithelium

Introduction

Transforming Growth Factor Beta 1 (TGFβ1) and other TGFβ family members (TGFβ2, TGFβ3) are members of a large superfamily of secreted cytokines and cognate receptors that regulate many aspects of cell and tissue function. The binding of a TGFβ ligand to a Type II TGFβ receptor (TGFβRII) dimer recruits and transphosphorylates a Type I TGFβ receptor (TGFβRI) dimer. The activated TGFβRI phosphorylates its downstream signaling mediators SMAD2 and SMAD3, transcription factors that form heteromers with SMAD4 to enter the nucleus and transcribe target genes (Massagué, 1998).

The TGFβ superfamily and receptors are critical in the development of skin and hair follicles, and in cycling of adult hair follicles. Most notably, bone morphogenic protein 4 (BMP4) signals for the development of hair follicle placodes from the surrounding interfollicular epidermis (IFE) during embryogenesis. However, TGFβ receptors and ligands also contribute strongly in the development and maintenance of hair and skin (Botchkarev et al., 1999). Neonatal hair follicles are strongly positive for TGFβRII in the outer root sheath (ORS), and TGFβRI is expressed broadly throughout the follicle and IFE. Expression of the TGFβ receptors fluctuate throughout the hair cycle with the strongest expression seen in late anagen and early catagen (Paus et al., 1997). Postnatal ablation of TGFβRII in the skin does not significantly affect overall skin morphology but does enhance sensitivity to tumorigenesis (Guasch et al., 2007). However, further analysis of adult mice with an epidermally inducible TGFβRII knockout showed that reentry into the hair cycle was delayed in the absence of TGFβ signaling (Oshimori and Fuchs, 2012). Blockage of TGFβRI activity produced a more severe phenotype than that of TGFβRII. Treatment with a TGFβRI inhibitor caused reduced hair follicle formation and delayed maturation in hair follicle grafts (Inoue et al., 2009), and extended anagen and delayed onset of catagen in adult mice (Naruse et al., 2017). Mouse embryonic stem cells treated with the same TGFβRI inhibitor showed reduced proliferation but no reduction in markers for stem cell pluripotency (Ogawa et al., 2007).

These studies reveal the importance of TGFβ family receptors in hair and skin development, however, comparatively little has been elucidated on the role of TGFβ ligands in these processes. TGFβ1 is expressed by the basal layer of the epidermis and the hair follicle ORS (Lyons and Moses, 1990; Glick et al., 1993; Foitzik et al., 2000). Expression of TGFβ1 is maximal in the ORS during late anagen and catagen and disappears in telogen (Foitzik et al., 2000). Whole-animal knockout TGFβ1−/− mice either die in utero, or at weaning age due to systemic inflammation and multi-organ failure (Shull et al., 1992; Kulkarni et al., 1993) thought to be caused by dysregulated T cell activation (Marie, Liggitt and Rudensky, 2006). These TGFβ1−/− mice have a slightly increased number of mature follicles at birth compared to wildtype (WT) littermates (Foitzik et al., 1999) and during the first postnatal hair cycle TGFβ1−/− mice exhibit prolonged early catagen and reduced entry into telogen at day 18 (Foitzik et al., 2000). Injection of TGFβ1 into the back skin (Foitzik et al., 2000) or overexpression of active TGFβ1 can block entry into anagen (Liu et al., 2001) but this may be due to suppression of keratinocyte proliferation.

In contrast to TGFβ1−/− mice, E19 TGFβ2−/− mice have fewer, less developed hair follicles compared to WT (Foitzik et al., 1999) and treatment of embryonic skin explants with TGFβ2 induced hair follicle development while TGFβ1 was suppressive (Foitzik et al., 1999). In adult mice injection of TGFβ2 specific or pan-TGFβ isoform neutralizing antibodies caused a reduction in hair follicle formation and maturation in hair follicle grafts (Inoue et al., 2009), while TGFβ2 signaling in the dermal papilla can relieve the inhibitory effect of BMP4 in hair follicle stem cells, and activate a new hair cycle (Oshimori and Fuchs, 2012). TGFβ2 expression by fibroblasts in the hair bulb is also linked to extracellular matrix remodeling and dermal fibroblast proliferation and differentiation during sustained Wnt/β-catenin induced ectopic hair follicle formation (Lichtenberger, Mastrogiannaki and Watt, 2016). Epidermal TGFβ2 expression was also shown to promote hair follicle formation by promotion of the transcriptional repressor B-lymphocyte-induced maturation protein 1 (BLIMP1) in the dermis (Telerman et al., 2017).

Taken together these results suggest that TGFβ signaling is essential for both hair follicle development and cycling. TGFβ2 appears to be critical for follicle development and initiation of a new hair cycle while TGFβ1 may be less important in follicle development but crucial for the catagen stage of the hair cycle. Keratinocytes have been demonstrated to be the primary source of TGFβ1 in the epidermis (Yang et al., 2019), although the consequence of keratinocyte-specific knockout of TGFβ1 on the physiology of the developing epidermis has not been directly demonstrated. Here we show that in newborn mice with targeted disruption of TGFβ1 in keratin 14 (K14) expressing tissues that both hair follicle development and differentiation of the epidermis are severally compromised, and this is associated with loss of p63 and LEF1 expression. Other squamous epithelia in the neonate are affected as well. TGFβ1 is also haploinsufficient, as mice heterozygous for TGFβ1 are runted with significant alterations in postnatal hair follicle morphogenesis. These studies reveal new insights into the complex role of TGFβ1 in development and differentiation of squamous epithelia.

Results

Targeted deletion of TGFβ1 in squamous epithelia causes developmental defects and perinatal lethality.

Keratin14-Cre (K14Cre)/Tgfβ1f/+ mice were crossed with Tgfβ1flox ex6/flox ex6 mice (Tgfβ1f/f) (Azhar et al., 2009) to delete TGFβ1 in the skin and other K14 expressing epithelia. Due to the orientation of the loxP sites in exon 6, Cre produces an inversion rather than deletion of exon 6 which can be identified with specific PCR primers (Fig. 1A, B). This inversion disrupts the Tgfβ1 reading frame leading to a null allele and as expected, TGFβ1 transcripts and protein were reduced in the basal layer in K14CreTgfβ1f/+, with greater loss in K14CreTgfβ1f/f mice (Fig. 1C, D). Transcript levels were 54% in K14CreTgfβ1f/+ compared to WT levels, and in K14CreTgfβ1f/f mice Tgfβ1 transcripts were 14% of that of WT. Protein levels of phospho-SMAD2/SMAD3 and total SMAD2/SMAD3 were also substantially reduced in newborn epidermis in K14CreTgfβ1f/+ and K14CreTgfβ1f/f mice (Fig. 1E), indicating loss of downstream TGFβ1 pathway signaling in the absence of TGFβ1 ligand.

Figure 1 – Characterization of TGFβ1 knockout.

Figure 1 –

Open in a new tab

(A) Schematic of loxp site locations (purple arrows) in the Tgfβ1 sequence showing inversion after Cre recombination. Primer sites for identification of this inversion by PCR are shown with black arrows (P1, P2 and P3). (B) Gel electrophoresis of products from PCR amplification using primers P1+P2, and P1+P3. P1+P2 distinguishes between floxed Tgfβ1 and WT sequence, and P1+P3 generates a unique product in mice with Cre-mediated inversion of Tgfβ1. (C) ΔΔCT analysis of Tgfβ1 transcripts in the newborn epidermis identified by qPCR, normalized to Gapdh (n=4 f/f, 2 Cre-f/+, 3 Cre-f/f). (D) Immunofluorescence with TGFβ1-specific antibody in newborn mouse skin. Keratin 14 (red) marks the boundary of the epidermal basal layer with the dermis; this boundary is indicated with a white line in the TGFβ1 (green) image. TGFβ1 positive regions are highlighted with white arrows. Areas of non-specific signal are seen in the stratum corneum, highlighted with yellow arrows; this signal is present in the absence of primary antibody. Scale bar = 20μm. (E) Western blot of newborn epidermal lysates probed with antibodies for phospho-SMAD2/SMAD3, total SMAD2/SMAD3 and Actin. Each lane represents the epidermis of one mouse. (F) ΔΔCT analysis of copy number of genes located of the same chromosome as floxed Tgfβ1, normalized to Actin. Genomic DNA was extracted from newborn mouse epidermis. Dashed line represents Tgfβ1f/f ΔΔCT and bars represent K14CreTgfβ1f/+ and K14CreTgfβ1f/f (n=5 f/f, 1 Cre-f/+, 3 Cre-f/f). (G) TUNEL assay showing infrequent apoptotic cells in the stratum corneum of newborn TGFβ1 knockout and wildtype skin. The basement membrane is indicated with a white line. Scale bar = 50μm.

Quantitative PCR analysis of epidermal genomic DNA for three genes present on the same chromosome as the floxed Tgfβ1 allele (Il16, Yif1b, and Cd22) showed no difference in copy number between genotypes (Fig. 1F), indicating no loss of that chromosome after Cre recombination. In addition, a TUNEL assay revealed no increase in apoptotic cells in K14CreTgfβ1f/+ or K14CreTgfβ1f/f mice compared to wild-type (Fig. 1G). Together these results indicate that phenotypes observed in K14CreTgfβ1f/f and K14CreTgfβ1f/+ mice are specifically due to the loss of functional exon 6 in Tgfβ1, and not due to chromosome loss and cell death caused by inversion induced chromosome abnormalities (Titen et al., 2020; Velegraki et al., 2021).

K14CreTgfβ1f/f mice were born with the expected Mendelian ratio but were absent at weaning (Fig. 2A). K14CreTgfβ1f/f newborns were smaller than their littermates (Fig. 2B) and died within 24 hours of birth with little-to-no milk in their stomachs, suggesting difficulty or a lack of nursing (Fig. 2B, black arrows). Over the course of nearly 2 hours, the newborn K14CreTgfβ1f/f mice lost about 1.25% of their starting weight but there was little change in the WT or hemizygous knockout genotypes (Fig. 2C). However there was no significant difference between the genotypes in toluidine blue dye uptake (Hardman et al., 1998) suggesting that perinatal death of mice lacking epidermal TGFβ1 was not due to a defect in epidermal barrier function (data not shown).

Figure 2 -. Loss of TGFβ1 in basal layer keratinocytes produces a neonatal lethal phenotype.

Figure 2 -

Open in a new tab

(A) Number of mice of each genotype born to and weaned from K14CreTgfβ1f/+ x Tgfβ1f/f parents. (B) Appearance and relative size of Tgfβ1f/f, K14CreTgfβ1f/+ and K14CreTgfβ1f/f neonates. Milk is visible inside the stomachs of Tgfβ1f/f and K14CreTgfβ1f/+ newborns, indicated by black arrows, but not in K14CreTgfβ1f/f littermates. Scale bars represent 5 mm. (C) Weight loss in neonatal pups during the first 150 minutes from birth. All samples represent measurements from at least three mice of each genotype. (D) Appearance of H&E-stained neonatal tongue cross-sections showing loss of filiform papillae in K14CreTgfβ1f/f. Bleeding is highlighted with a black arrow. Scale bars represent 20 μm. (E) Quantification of tongue mucosa thickness (measured from basement membrane to top of filiform papillae) by genotype (n=10 f/f, 10 Cre-f/+, 9 Cre-f/f). Data points represent average of at least 20 measurements taken for each mouse. (F) H&E stained neonatal esophagus cross-sections. (G) Quantification of esophagus mucosa width by genotype (n=3 f/f, 4 Cre-f/+, 3 Cre-f/f). Data points represent average of at least 15 measurements taken for each mouse. Scale bars represent 20 μm.

The lingual mucosa of K14CreTgfβ1f/f mice was thinner than both WT and K14CreTgfβ1f/+, and lacked normal filiform papillae (Fig. 2D, E). Additionally, there was mild blistering at the lingual basement membrane and increased vascularity that was not found other genotypes. The filiform papillae of K14CreTgfβ1f/+ tongues were not as prominent as in WT, but the thickness of the epithelium was not significantly different. However, the esophageal mucosa was significantly thinner in both K14CreTgfβ1f/f and K14CreTgfβ1f/+ than in WT littermates (Fig. 2F, G). Together these results suggest that oral and esophageal mucosa is thinned in K14CreTgfβ1f/f pups to the point that normal nursing is hindered, which could be a major contributor to the neonatal mortality in K14CreTgfβ1f/f mice.

In addition to aberrant morphology of the oral epithelium there was a significant defect in epidermal development in mice lacking epidermal TGFβ1. The epidermis of newborn K14CreTgfβ1f/f mice was significantly thinner than that of K14CreTgfβ1f/+ (12 μm vs 24 μm) and WT (30 μm) mice and had 50 percent fewer hair follicles compared to K14CreTgfβ1f/+ and WT mice (Fig. 3A–C). There was also a significant difference in hair follicle developmental stage in the absence of epidermal TGFβ1 (Fig. 3D). As expected, the majority of hair follicles in newborn control mice were in stage 4–6 (intermediate) of morphological development, while 35% were at stages 7–8 and 5% at Stage 1–3 (Paus et al., 1999). In contrast, the majority of follicles in K14CreTgfβ1f/f mice were at stages 1–3, while only 30% were at intermediate stages and 5% at stages 7–8. K14CreTgfβ1f/+ mice were midway between WT and full deletion with 60% of follicles in intermediate stages, but 20% at stages 1–3 and 20% at stages 7–8. Thus depletion or reduction of TGFβ1 levels in epidermal keratinocytes retards normal hair follicle development.

Figure 3 -. Loss of TGFβ1 in the newborn epidermis causes reduced hair follicle numbers and retards development.

Figure 3 -

Open in a new tab

(A) H&E stained back skin sections taken from newborn Tgfβ1f/f, K14CreTgfβ1f/+ and K14CreTgfβ1f/f littermates. Scale bar = 100 μm. (B) Quantification of number of hair follicles by genotype. (n=8 f/f, 5 Cre-f/+, 7 Cre-f/f mice). The average skin length that follicles were counted from was 12000 μm, and the shortest skin quantified was 4000 μm. (C) Quantification of interfollicular epidermis thickness by genotype. (n=8 f/f, 5 Cre-f/+, 5 Cre-f/f mice). (D) Classification of newborn hair follicles by stage, grouped into early, intermediate and late stages. Each column represents one mouse. (n=7 f/f, 4 Cre-f/+, 6 Cre-f/f mice).

Partial deletion of TGFβ1 leads to postnatal runting and abnormal hair follicle morphogenesis.

Although no perinatal lethality occurred with K14CreTgfβ1f/+ mice, severe but variable runting was evident during postnatal development. By p21 K14CreTgfβ1f/+ mice weighed 20% to 30% less than their WT littermates (Fig. 4A) and had sparse coats and flaky skin. (4B, C). Surprisingly, while there were roughly equal numbers of male and female K14CreTgfβ1f/+ mice at birth as determined using sex chromosome specific PCR (McFarlane et al., 2013), at weaning 82% (29/35 weaned K14CreTgfβ1f/+ mice over 4 years) were male, while only 17% (6/35) were female, suggesting selective mortality of female K14CreTgfβ1f/+ mice through an unknown mechanism.

Figure 4 – Hemizygous deletion of Tgfβ1 causes runting and sparse hair coat during postnatal development.

Figure 4 –

Open in a new tab

(A) Graph of mouse weights from birth to weaning. Average n is 8, all data points have an n of at least 3. (B) Appearance of K14CreTgfβ1f/+ skin and coat at 18 days old. (C) Appearance of Tgfβ1f/f and K14CreTgfβ1f/+ littermates at 21 days old. (D) All hair types are present in K14CreTgfβ1f/+ mice at 21 days old, but hair shafts have a darker appearance with disorganized medulla, or even absent medulla.

All four hair types (guard, awl, auchene and zigzag) were present in the expected ratios in K14CreTgfβ1f/+ mice, indicating that the thinner coat was not due to loss of one type of hair. However, the K14CreTgfβ1f/+ hair shafts were generally darker in appearance than WT and the medulla columns were more disorganized (Fig. 4D). Occasionally, hair shafts were observed with no medulla columns in K14CreTgfβ1f/+ mice; these hairs could be either with or without pigmentation.

Given that TGFβ1 has a well-established role in regulation of the hair cycle, particularly in the initiation of catagen (Foitzik et al., 2000), we examined the effect of partial epidermal TGFβ1 knockout on hair morphogenesis. No significant difference was seen at p5 in hair follicle density or maturation stage (Fig. 5A), but by p9 all hair follicles in WT mice were fully mature and in late anagen, with the hair bulb at the base of the adipose layer and a prominent hair shaft rising from the follicle, as expected (Paus et al., 1998). In contrast, immature follicles can be seen in p9 K14CreTgfβ1f/+ mice (highlighted by black arrows) suggesting a delay in hair follicle maturation. The follicles in p9 K14CreTgfβ1f/+ had a twisted disorganized morphology, with the hair bulb of adjacent follicles frequently oriented in different directions (Fig. 5A, B).

Figure 5 – Hair follicle morphogenesis is disrupted when TGFβ1 is depleted in the basal layer.

Figure 5 –

Open in a new tab

(A) H&E stained back skin sections harvested at timepoints from p5 to p28. Immature hair follicles in p9 are highlighted with black arrows. Scale bars represent 100μm. (B) Hyperkeratotic hair follicles (highlighted with yellow arrows) are seen in p9, p18 and p21 H&E stained skin sections from K14CreTgfβ1f/+ mice. (C) One instance of extensive inflammation and epidermal hyperkeratosis observed at p18.

At p17, the majority of WT follicles were in early catagen, while in K14CreTgfβ1f/+ mice the hair follicles were in different stages, with some remaining in anagen and others appeared to be in later stages of catagen (catagen V). At p18, WT follicles were in later stages of catagen with most follicles regressed from the adipose layer into the dermis. In K14CreTgfβ1f/+ mice the follicle bulbs remained deeper in the adipose layer. In a few K14CreTgfβ1f/+ mice the epidermis was thickened and hyperkeratotic, the hair follicles were enlarged (Fig. 5B), and extensive inflammatory infiltrate was seen in the dermis (Fig. 5C). At p19 and p21 most WT follicles were in late catagen/telogen. While some p19 K14CreTgfβ1f/+ follicles appeared to have a telogen morphology, many were elongated with hair bulbs that remained at the border between the adipose and muscle layer, although they were thinner and did not have an anagen morphology. The hair follicles of p21 K14CreTgfβ1f/+ mice were highly irregular in their orientation and appearance; follicles with anagen, catagen and telogen morphology were observed in close proximity to each other. The sebaceous glands were also abnormally large in these skins. Enlarged, hyperkeratotic and inflamed hair follicles were occasionally found in p21 K14CreTgfβ1f/+ skins, and less frequently at earlier timepoints (Fig. 5B, highlighted with yellow arrows).

At p24 all WT follicles were in telogen, while K14CreTgfβ1f/+ follicles were in late anagen based on the position of the hair bulb at the border between the adipose and muscle layer and thin dermal papillae. By p28, WT follicles were entering anagen while K14CreTgfβ1f/+ follicles were in late anagen, although the follicle morphology was aberrant. Together, this data shows that the level of TGFβ1 expression is critical for normal hair cycling. Hemizygous deletion of TGFβ1 in the basal layer and hair follicle appears to disrupt the normal onset of catagen and prevent telogen of the first hair cycle, as well as disrupt normal hair follicle morphology and orientation.

Mice with reduced TGFβ1 in basal keratinocytes have abnormal stratification, and impaired basement membrane and cell-cell adhesion.

To determine the basis for the thinned epithelium we examined proliferation and differentiation markers in the epidermis. We found that proliferation of keratinocytes in the basal layer was significantly downregulated in newborn K14CreTgfβ1f/f and K14CreTgfβ1f/+ mice (Fig. 6A,B). There were 78% fewer Ki67 positive interfollicular keratinocytes in K14CreTgfβ1f/f compared to WT and 31% fewer than WT in K14CreTgfβ1f/+ epidermis. As previously reported, there was no detectable difference in apoptosis between genotypes as measured by a TUNEL assay (Fig. 1G).

Figure 6 – TGFβ1 depleted epithelium has dysregulated proliferation, stratification and a disrupted basement membrane.

Figure 6 –

Open in a new tab

(A) Immunofluorescence of keratin 14 (red) and Ki67 (green) in newborn skin frozen sections. Scale bar = 25 μm. The boundary between the interfollicular epidermis (IFE) basal layer and the dermis is indicated with a white line. (B) Number of Ki67 positive epidermal keratinocytes per 1000 pixels2 of basal layer by genotype (n= 3 f/f, 3 Cre-f/+, 3 Cre-f/f). Points represent average number of positive cells found in three images per mouse. (C) Immunofluorescence (IF) of keratin 5 (red) and keratin 1 (green) in newborn skin. Scale bar = 50 μm. Basal keratinocytes with K1 expression are indicated with white arrows, and regions where the basal layer is two cells thick are highlighted with white brackets. (D) IF of K5 (red) and loricrin (green). (E) Immunohistochemical staining for Integrin α6 in newborn skin. Scale bar = 50 μm (F) IF imaging of K14 (red) and E-cadherin (green). The boundary between epidermis and dermis, as identified by K14 expression, is shown with a white dashed line. (G) K14 and (H) K1 staining in hard palate of newborn oral cavity. Scale bar = 20μm. (I) BLIMP1 (green) and K14 (red) immunofluorescence in newborn skin. Granular keratinocytes double-positive for K14 and BLIMP1 are indicated with yellow arrows. Scale bar = 50 μm.

The basal layer marker Keratin 5 (K5); the spinous layer differentiation marker Keratin 1 (K1) (Fig. 6C); and the granular layer marker loricrin (Fig. 6D) were present in all genotypes, but expression of the suprabasal differentiation markers was reduced in mice lacking epidermal TGFβ1. In addition, the layer of K5+K1− cells was consistently one cell layer thick in WT, whereas in K14CreTgfβ1f/f epidermis there were regions where this layer widens to two cells (white brackets, Fig 6C). Similarly, a very small number of WT basal keratinocytes were positive for both K5 and K1, indicated by white arrows in Fig. 6C, which are presumed to represent cells in the process of differentiating. These transitional cells are not found in K14CreTgfβ1f/f epidermis.

Although there was no gross blistering of the skin in newborn K14CreTgfβ1f/f mice, in skin sections there was localized detachment of the epidermis at the junction of the basal layer and first suprabasal layer or basement membrane, suggesting some degree of skin fragility. As expected, (Sonnenberg et al., 1991) expression of the hemidesmosome component integrin alpha 6 (ITGα6) was seen on the plasma membrane of keratinocytes in the basal layer of WT and K14CreTgfβ1f/+ mice (Fig. 6E). However, in K14CreTgfβ1f/f newborns expression was focally extended into the suprabasal layers and pockets of ITGα6 negative cells were found in the basal layer. Additionally, the epithelial adherens junction component E-cadherin was expressed strongly throughout the epidermis of WT newborn skin, but expression was reduced in K14CreTgfβ1f/f and basal keratinocytes negative for E-cadherin were observed (Fig. 6F). This suggests a loss of epithelial identity in these cells and compromised cell-cell signaling and connection. Conversely, expression of K14 appears to persist beyond the basal layer and it remains present in suprabasal cells. Together, these results indicate disruption of the normally strict separation of basal and suprabasal layers and compromised structural integrity of the epithelium.

Similar to the epidermis, squamous differentiation in the oral cavity was also disrupted. In the hard palate there was a reduction of stratified layers and substantially reduced expression of K1 (Fig. 6H), although K14 was normal (Fig. 6G). This thinning of the oral epithelium may also contribute to the morbidity of K14CreTgfβ1f/f newborns.

Since there were fewer loricrin positive cells in the granular layer (Fig. 6D), we examined expression of BLIMP1, a transcriptional repressor required for terminal differentiation (Fig. 6I). Consistent with published data (Lesko et al., 2013), in WT mice BLIMP1 positive cells are found in the dermal papillae, the sebaceous bulb and the granular layer of the epidermis, while the spinous layer keratinocytes were not positive for either BLIMP1 or K14. Little-to-no sebaceous gland expression was found in K14CreTgfβ1f/f, which may reflect the previously observed delay in hair follicle maturation as the sebaceous gland does not develop until later morphological stages. In the granular layer, there was a similar number of BLIMP1 positive cells and level of BLIMP1 expression relative to WT mice but there was no layer of negative cells between the basal (K14+) and granular (BLIMP1+) layers. We also observed keratinocytes positive for both K14 and BLIMP1 in K14CreTgfβ1f/f mice (indicated with yellow arrows). This data suggests that stratification is occurring in K14CreTgfβ1f/f and all expected layers are present, but each layer is thinner and the distinction between layers is blurred.

Reduced p63 and LEF1 expression in epidermis of mice lacking TGFβ1.

TGFβ1 is a potent inhibitor of epithelial proliferation, suggesting that the phenotype caused by its absence in the epidermis is mediated indirectly through other signaling pathways. We examined expression of two critical transcription factors, p63 and LEF1, to understand the mechanism underlying the reduced proliferation and altered differentiation in the newborn epidermis. Tp63 is the master transcription factor controlling proliferation and differentiation in the basal layer of stratified squamous epithelium (Parsa et al., 1999), primarily by its ΔNp63 isoform (Yang et al., 1998). Using both a pan-isoform targeting p63 antibody and a ΔNp63 isoform specific antibody, we found no significant difference between the number of p63 positive cells in K14CreTgfβ1f/+ and WT IFE. In contrast there were significantly fewer total p63 and ΔNp63 positive cells in the K14CreTgfβ1f/f basal layer (Fig. 7A,B quantitated in 7D,E). There was a significant downregulation of ΔNp63 transcripts between WT and K14CreTgfβ1f/+, and between WT and K14CreTgfβ1f/f (Fig 7F). The TAp63 isoform is regulated by a different promoter from the ΔNp63 isoform and is expressed at vastly lower levels in keratinocytes than the ΔNp63. Although we were unable to detect the TAp63 isoform in the skin by immunofluorescence, Tap63 transcripts were also significantly downregulated in K14CreTgfβ1f/f compared to WT epidermis (Fig. 7G).

Figure 7 – TGFβ1 is required for basal keratinocyte identity and epidermal Wnt signaling.

Figure 7 –

Open in a new tab

(A) p63 (green) and K14 (red) immunofluorescence in newborn skin. Scale bar = 25 μm. (B) ΔNp63 (green) and K14 (red) immunofluorescence in newborn skin. Scale bar = 25 μm. (C) Immunofluorescence of LEF1 (green) in newborn skin. Scale bar = 25 μm. The boundary between IFE and dermis is indicated with a white line, based on K14 expression. IFE areas with LEF1 positive cells are highlighted with a yellow arrow. (D) Quantification of p63 positive cells per 1000 pixels2 of basal layer measured (n=3 f/f, 2 Cre-f/+, 3 Cre-f/f mice). (E) Quantification of ΔNp63 positive cells per 1000 pixels2 of basal layer measured (n=3 f/f, 3 Cre-f/+, 3 Cre-f/f mice). (F) Fold change in epidermal ΔNp63 transcripts. (n=3 f/f, 2 Cre-f/+, 3 Cre-f/f mice). (G) Fold change in epidermal TAp63 transcripts. (n=3 f/f, 2 Cre-f/+, 3 Cre-f/f mice).

LEF1 is a transcription factor activated by β-catenin as part of the Wnt signaling pathway, and upregulated by TGFβ signaling (Cordray and Satterwhite, 2005), with an essential role in hair follicle development (Zhou et al., 1995). In WT skins, LEF1 was expressed in the hair follicle bulb, dermal fibroblasts and weakly in basal keratinocytes (Fig. 7C), matching the expected localization (Phan et al., 2020). However, there were fewer LEF1 positive cells in all three locations in K14CreTgfβ1f/+, with stronger downregulation in K14CreTgfβ1f/f. Together, these results suggest that TGFβ1 signaling is critical for expression of both p63 and LEF1, and the absence of TGFβ1 leads to both loss of proliferation and aberrant epidermal follicular development in newborn mice.

Homeostasis of the interfollicular epidermis is also disrupted in mice with hemizygous deletion of TGFβ1

In addition to affecting the hair cycle, hemizygous deletion of Tgfβ1 significantly altered hair cycle dependent changes in homeostasis of the IFE. The IFE of K14CreTgfβ1f/+ mice was significantly thinner than that of WT mice at birth (Fig. 3C), although by p5 this difference was lost with the normal thinning of the WT epidermis (Fig. 5A, quantified in Fig. 8A). From p9 to p21 the IFE of K14CreTgfβ1f/+ mice exhibited a cyclical expansion to twice the thickness of WT, followed by thinning.

Figure 8 – TGFβ1-ablated interfollicular epidermis displays hair cycle-linked changes in thickness, as well as dysregulated proliferation and p63 expression.

Figure 8 –

Open in a new tab

(A) Epidermal thickness in Tgfβ1f/f and K14CreTgfβ1f/+ skin in p5, p9, p19 and p21. Average n is 6 mice; all samples represent n of at least 3 mice. (B) Immunofluorescence images of p5, p9 and p21 Tgfβ1f/f and K14CreTgfβ1f/+ skin with Ki67 (green) and keratin 14 (red). Scale bars represent 50 μm (C) Quantification of percentage of Ki67 cells in epidermal basal layer. All samples represent 3 mice. (D) p63 immunofluorescence and (E) quantification of p63 positive cells per 1000 pixels2 measured of epidermis. All samples represent 3 mice. Scale bars represent 50 μm. (F) Immunofluorescence of keratin 14 (red) overlaid with keratin 10 (green) in p5 and p9 WT and K14CreTgfβ1f/+ mice. Cells double-positive for K14 and K10 are highlighted with white arrows. Scale bars represent 50 μm. (G) Immunofluorescence of keratin 14 (red) overlaid with loricrin (green) in p5 and p9 WT and K14CreTgfβ1f/+ mice. Nucleated loricrin-positive granular keratinocytes are highlighted with yellow arrows. Scale bars represent 50 μm.

Although there was no difference in the number of Ki67 positive cells at p5, by p9 there were at least twice as many proliferating cells in K14CreTgfβ1f/+ epidermis, paralleling the increased epidermal thickness at this timepoint (Fig. 8A–C). Proliferating cells were sparse in both WT and K14CreTgfβ1f/+ during catagen in p17. Surprisingly, the unusually thick IFE of p21 K14CreTgfβ1f/+ displayed the same percentage of proliferating cells as the thinner WT epidermis. This result suggests that additional factors are involved with the thickness of the K14CreTgfβ1f/+ IFE besides cell proliferation.

Based on the observed dysregulation of keratinocyte proliferation and differentiation we examined expression of p63, since it had been altered in the epidermis of newborn K14CreTgfβ1f/f mice. p63 was expressed predominantly in the basal layer and infrequently in the spinous layer in both p5 WT and K14CreTgfβ1f/+ (Fig. 8D, with quantification in 8E), although more frequent suprabasal expression was seen in K14CreTgfβ1f/+. Significantly altered expression was found in p9 K14CreTgfβ1f/+, where there were 40% more p63 positive cells and expression was found in 2–3 layers above the basal layer. This pattern of increased suprabasal expression was also observed in p17 and p21, and there were significantly more p63 positive cells at p21 than in WT. This result shows that expansion of the p63 positive population is not solely dependent on increased basal layer proliferation, but may be linked to the aberrant expansion of the basal layer and altered squamous differentiation.

An expansion of the K14 positive cell layer was also found in postnatal K14CreTgfβ1f/+ mice. Although K10 was similarly expressed in WT and K14CreTgfβ1f/+ mice at p5 and p9 (Fig. 8F) the layer of K14+K10− keratinocytes in p9 K14CreTgfβ1f/+ mice had expanded in multiple regions to 2–3 cell layers thick compared to the strict one cell layer always found in normal epidermis (highlighted by a white bracket). Additionally, infrequent but easily detectable K14+K10+ keratinocytes were present in the basal layer of WT epidermis (Fig. 8F, white arrows) but these were rarely found in K14CreTgfβ1f/+ mice, suggesting a disruption in onset of normal differentiation. The loricrin-expressing granular layer was as thick in the K14CreTgfβ1f/+ epidermis as WT at p5 (Fig. 8G) but did not decrease in thickness as much as WT by p9. The number of nucleated loricrin-expressing keratinocytes in the granular layer of K14CreTgfβ1f/+ mice was increased compared to that of WT mice in both p5 and p9 (Fig. 8G, yellow arrows) again suggesting dysregulation of the normal rate of differentiation. Collectively, this data shows that although all the expected stratified squamous epithelial layers are present in K14CreTgfβ1f/+ mice, squamous differentiation is abnormal and the distinction between layers less defined. Combined with the data for proliferation and p63, these results indicate a role for TGFβ1 in the maintenance of stratified squamous epithelium in older mice independent of its role in epithelial development.

Keratinocyte TGFβ1 controls epidermal melanocyte population and migration.

Although the major impact of reduced or absent TGFβ1 was on keratinocytes in the epidermis we also observed significant alterations in melanocyte biology. At birth there was hyperpigmentation on the pawpads of K14CreTgfβ1f/f and K14CreTgfβ1f/+ pups (Fig. 9A), and in the paw pads (Fig. 9B) and skin (Fig. 9D) of postnatal K14CreTgfβ1f/+ mice, although it was lost by 9-week old hemizygous mice (Fig. 9C). A Fontana-Masson stain confirmed that the hyperpigmentation was excessive melanin, and revealed increased melanin in the hair shaft, follicle sheath and bulb, and IFE of p1 and p9 TGFβ1 homozygous and hemizygous knockout mice, with the highest staining seen at p9 (Fig. 9E). In contrast, WT mice only show strong melanin staining in the hair shaft, and sparse staining in the hair follicle bulb. By p21, K14CreTgfβ1f/+ mice had little IFE localized melanin, but excessive melanin was found in the hair follicle sheath. High levels of melanin were also observed in the volar pad epidermis of K14CreTgfβ1f/+ mice at p9, and this is reduced but still present by p21 (Fig. 9F).

Figure 9 – Unusually high levels of melanin are found in the skin and hair follicles of TGFβ1 knockout mice.

Figure 9 –

Open in a new tab

(A) Dark pigmentation is present on the paws of newborn K14CreTgfβ1f/+ and K14CreTgfβ1f/f mice, but not WT littermates. (B) Dark pigmentation is seen on paw pads of p21 K14CreTgfβ1f/+ mice, but not on the paws of (C) 9 week old mice. (D) Skin appearance of p21 Tgfβ1f/f and K14CreTgfβ1f/+ littermates after hair removal. (E) Skins from p1, p9 and p21 Tgfβ1f/f and K14CreTgfβ1f/+ mice treated with a Fontana Masson stain to highlight melanin deposits in black, and counterstained with Nuclear Fast Red. (F) Paw pad cross sections from p9 and p21 stained with Fontana Masson and Nuclear Fast Red. (G) Immunofluorescence of cKIT (green) and Keratin 14 (red) on back skin sections.

One factor required for melanocyte survival and migration, cKIT (Kunisada et al., 1998; Botchkareva et al., 2001), is inhibited by TGFβ (Dubois et al., 1994; Yang et al., 2008) and we hypothesized that loss of epidermal TGFβ1 was enabling expansion of melanocytes in the IFE. Melanocytes are normally found first in the embryonic IFE, and these cells migrate to the root sheath of immature hair follicles before localizing in the matured hair bulb by p9 (Peters et al., 2002). Using an antibody for cKIT, we found that this pattern of expression was present in WT newborn skin, but K14CreTgfβ1f/f skins showed more cKIT positive cells in the IFE and follicle sheath (Fig. 9G). By p9, no cKIT positive cells remained in the IFE of either genotype, but K14CreTgfβ1f/+ mice still had a high number of follicular melanocytes. This suggests that reduced levels of TGFβ1 in the epidermis leads to an increase in the melanocyte population and delayed melanocyte migration, and subsequent visible hyperpigmentation.

Discussion

We generated mice with knockout of TGFβ1 targeted to K14 expressing tissue. These mice were found to have substantial loss of both phosphorylated and total SMAD2/3, indicating repression of downstream TGFβ1 signaling in these mice. Loss of total SMAD2 and SMAD3 at both the transcript and protein level has been reported after siRNA inhibition of TGFβRI and/or TGFβRII (Wang et al., 2017), which suggests that TGFβ signaling is required for SMAD2/3 expression. Newborn mice homozygous for ablated TGFβ1 showed perinatal lethality and substantially disrupted stratified squamous epithelium in the skin, esophagus and hard palate, and most severely in the tongue. Hemizygous mice survived to adulthood but displayed runted growth, visibly thin coats and unusually dark pigmentation. The esophageal mucosa of K14CreTgfβ1f/+ newborns showed slight, but significant, thinning and this may lead to a reduced intake of milk and subsequent slower growth prior to weaning age. Together these results show that TGFβ1 is critical for normal development and differentiation of multiple squamous epithelia as well as postnatal morphogenesis of the epidermis. Furthermore, these results indicate that paracrine TGFβ1 from nonepithelial sources, or other TGFβ isoforms which are expressed by these tissues cannot compensate for either complete or partial reduction in autocrine TGFβ1 production by basal keratinocytes in squamous epithelia. While the Tgfβ1flox ex6 allele used in this study had been hypothesized to cause post-recombination damage to chromosome 7 (Titen et al., 2020; Velegraki et al., 2021), we observed no loss of chromosome 7 and no increase in apoptotic cells in TGFβ1 knockout pups, indicating that the observed phenotype is a direct consequence of TGFβ1 ablation.

Our results provide compelling evidence for distinct roles for TGFβ1 in development of stratified epithelium compared to postnatal morphogenesis, most strikingly in the epidermis. Although we did not directly measure the reduction in epidermal TGFβ1 protein in K14CreTgfβ1f/+ mice relative to WT mice, our results demonstrate that reduction, but not complete loss, of TGFβ1 can have profound effects on both keratinocytes and melanocytes in the epidermis. Newborn K14CreTgfβ1f/f mice displayed altered IFE with thinner K1 and loricrin positive layers, reduced proliferation, and persistence of K14/K5 expression beyond the basal layer, to the extent that some suprabasal keratinocytes were positive for both K14 and BLIMP1. This persistence of K14 may be due to a compensation effect where K14 filaments are stabilized and retained in cells with reduced K10 expression (Reichelt et al., 2001).

We also observed hyperpigmentation in TGFβ1 knockout tissue and found that this correlated with an increased number of cells positive for the melanocyte marker cKIT, and as well as delayed migration of these cells to the hair follicles. An additional source of this hyperpigmentation may be due to TGFβ1-mediated inhibition of enzymes for melanin synthesis in melanocytes (Kim, Park and Park, 2004; Murakami, Matsuzaki and Funaba, 2009; Nishimura et al., 2010); therefore the increase in melanin could be due to loss of this inhibition. Additionally, mice with systemic overexpression of hepatocyte growth factor (HGF) display a similar hyperpigmented phenotype (Takayama et al., 1996) as our epidermal TGFβ1 knockout mice. HGF has been shown to be inhibited by the TGFβ pathway in fibroblasts (McKeown et al., 2003; Yi et al., 2014) and therefore crosstalk between epidermal TGFβ1 and dermal HGF may be required to ensure appropriate localization of melanocytes in the skin.

Most studies have linked activation of the TGFβ receptors (Paus et al., 1997; Inoue et al., 2009; Oshimori and Fuchs, 2012) and TGFβ2 (Foitzik et al., 1999; Inoue et al., 2009; Oshimori and Fuchs, 2012) with hair follicle formation but TGFβ1 overexpression with suppression of hair follicle formation and proliferation (Sellheyer et al., 1993; Liu et al., 2001) and conventional knockout of TGFβ1 with increased hair follicle number and proliferation (Foitzik et al., 2000). In contrast, our results show significant suppression of hair follicle formation and development in newborn mice with a targeted deletion of TGFβ1 in epidermal keratinocytes. This hair follicle phenotype is similar to mice with pharmacological inhibition of TGFβRI, which showed reduced hair follicle development and maturation (Inoue et al., 2009) and delayed catagen (Naruse et al., 2017). It is possible that supraphysiological expression of active TGFβ1 provides a strong growth inhibitory signal that overrides normal inductive processes in hair follicle development. Conversely, increased follicle numbers in the skin of TGFβ1 null mice could reflect the net effect of loss of TGFβ1 in keratinocytes as well as other cell types such as dermal papillae cells. The reduction in expression of the key epidermal and hair follicle regulators p63 and Lef1 strengthens the idea that keratinocyte TGFβ1 may have important effects on hair follicle development distinct from its well characterized anti-proliferative activity.

Consistent with multiple studies (Paus et al., 1997; Foitzik et al., 2000), catagen and telogen of the first hair cycle in K14CreTgfβ1f/+ mice were delayed or absent in many follicles with rapid entry into anagen. Thus, autocrine TGFβ1 is essential for normal regulation of the hair cycle. The disorganized growth of follicles in the dermis also indicates that keratinocyte TGFβ1 is important in downgrowth and orientation of follicles during morphogenesis. In contrast to neonatal epidermis of K14CreTgfβ1f/f mice, the IFE of K14CreTgfβ1f/+ mice was thickened and hyperproliferative at p9, which is consistent with loss of anti-proliferative effects of TGFβ1. After thinning at p17, the IFE thickened again at p21 but with no increase in proliferation. This thickening without proliferation is similar to the phenotype of conditional K14-targeted TGFβ1 knockout in adult mice, where significant epidermal acanthosis was also observed (Yang et al., 2019). The fluctuation in IFE thickness in K14CreTgfβ1f/+ mice mirrors the cyclical expression of TGFβ1 and its receptors peaking in late anagen and catagen (Paus et al., 1997; Foitzik et al., 2000) suggesting direct TGFβ1 involvement in this process. TGFβ1 ligand has also been shown to suppress the growth and differentiation of sebaceous glands (McNairn et al., 2013), and therefore the enlarged sebaceous glands in p21 K14CreTgfβ1f/+ mice appear to be caused by loss of this regulation. Additionally, an inflammatory infiltrate was occasionally observed around K14CreTgfβ1f/+ hair follicles, and this was likely caused by loss of anti-inflammatory TGFβ1 signaling (Yoshimura, Wakabayashi and Mori, 2010). It is possible that hyperkeratosis in hair follicles or increased inflammation in the dermis could damage the follicle or prevent normal hair shaft formation leading to sparseness of the coat in K14CreTgfβ1f/+ mice. Similar effects of inflammation on hair follicles were observed in mice with deletion of the EGFR (Hansen et al., 1997).

TGFβ signaling has been shown to strongly upregulate expression of Wnt pathway transcription factor LEF1 in a SMAD-independent manner (Cordray and Satterwhite, 2005). Lef1 transcripts are also degraded by miR-203 (Thatcher et al., 2008), a microRNA normally downregulated by TGFβ (Ding et al., 2013), and therefore reduction of TGFβ may lead to accumulation of miR-203 and excessive degradation of Lef1. Complementary to these findings, we found fewer LEF1 positive cells in the hair follicles, fibroblasts and IFE of K14CreTgfβ1f/f mice. As whole-animal knockout of Lef1 was shown to result in thin skin with fewer, poorly developed hair follicles (van Genderen et al., 1994) it therefore appears likely that this downregulation of LEF1 contributes to the neonatal skin and hair follicle phenotype in our TGFβ1 knockout mice. However, by p21 LEF1 expression is confined to the hair bulb and sebaceous gland (Phan et al., 2020), and we observed this to also be the case in p5, p9 and p21 WT and K14CreTgfβ1f/+ mice (data not shown). This would therefore suggest that LEF1 is not involved in the dysregulated epithelial homeostasis and hair cycle changes found in older mice, and a different molecular mechanism is responsible for the phenotype of mice with reduced TGFβ1 after p1.

Given the critical role of p63 in development of stratified squamous epithelia (Mills et al., 1999; Yang et al., 1999), its importance in regulation of epidermal progenitor cells and keratinocyte proliferation (Senoo et al., 2007) as well as stratification (Koster et al., 2004; Truong et al., 2006), it is possible that both reduction and increase in p63 expression mediates some of the phenotypic effects of partial or complete epidermal TGFβ1 deletion. The loss of p63 in basal keratinocytes also leads to a more migratory phenotype and increased EMT (Barbieri et al., 2006), as well as basement membrane disruption (Koster et al., 2007) which could explain the disorganized, irregular appearance of the basal layer in K14CreTgfβ1f/f and K14CreTgfβ1f/+ skins, and increased fragility and aberrant α6 integrin and E-cadherin expression of K14CreTgfβ1f/f epidermis. In the epidermis of newborn K14CreTgfβ1f/f mice the number of p63 positive basal keratinocytes was significantly reduced, and given the role of p63 in keratinocyte proliferation (Senoo et al., 2007) it is possible that the loss of p63 directly leads to the loss of basal layer proliferation in these skins. In p5–p21 K14CreTgfβ1f/+ mice there was an unusual expansion of p63 positive keratinocytes above the basal layer, which may contribute to the thickened epidermis seen in these mice.

Previous studies have shown that TGFβ1 can either induce or suppress p63 expression depending on the context. In cancer cell lines TGFβ1 can directly induce p63 expression (Fukunishi et al., 2010; Sundqvist et al., 2020), and depending on cellular context can either promote (Bui et al., 2020) or repress (Ding et al., 2013) microRNAs that degrade ΔNp63 transcripts (Lena et al., 2008). Although it remains to be determined, it is possible that depending on the context reduced or absent TGFβ1 can lead to both reduced or elevated p63 levels.

In conclusion, our study shows that autocrine TGFβ1 has an essential role in the development of stratified squamous epithelium, and in the maintenance of these tissues after birth. Our results point to effects that are direct action of TGFβ1 as well as indirectly through p63 and LEF1. These results have significant implications for understanding human diseases and mouse models associated with alterations in TGFβ1 signaling or expression.

Materials and Methods

Generation of mice with epidermal deletion of TGFβ1

All studies were performed in compliance with U.S. Department of Health and Human Services Guide for the Care and Use of Laboratory Animals and after approval by The Pennsylvania State University Institutional Animal Care and Use Committee. KRT14cre1Amc/J mice (K14Cre) (Dassule et al., 2000) were obtained from Jackson labs and crossed with Tgfβ1flox ex6 mice (Tgfβ1f/f) (Azhar et al., 2009) that had been backcrossed 6 generations onto C57BL/6J mice to create K14CreTgfβ1f/+ mice. K14CreTgfβ1f/+ mice were crossed with Tgfβ1f/f mice to create Tgfβ1Δep mice. Due to the orientation of the loxP sites in exon 6, the action of Cre recombinase produces an inversion rather than deletion, although nonetheless inactivating the Tgfβ1 locus. Both male and female mice were used in experiments.

PCR genotyping

Tail snips of <1 mm were excised from pups at weaning, or ear snips were taken from harvested adult animals. Biopsies were boiled in 50 mM NaOH with 0.2mM EDTA for 60 minutes then neutralized in 90 mM Tris-Cl (pH 8.0). DNA was precipitated from the supernatant with 50% v/v isopropanol, pelleted and resuspended in 10mM Tris-HCl (pH 7.4) with 0.1mM EDTA. The following PCR primers were used to validate inversion of Tgfβ1 exon 6 in either the targeted or conditional mice: exon 6 (P1) 5’ GGGCTACCATGCCCACT; intron 6–7 (P2) 5’CTTCTCCGTTTCTCTGTCACCCTAT; inversion intron 5–6 (P3) 5’ CCACCCAGGGAGCTTTAACTC (Azhar et al., 2009) using 1 μL genomic DNA per 20 μL reaction. P1+P2 produces a 182 bp product with unfloxed Tgfβ1 allele and 220 bp product with floxed Tgfβ1 allele. P1+P3 produces a 370 bp product from the inverted Tgfβ1fl allele only in the presence of Cre and no product with WT or unrecombined Tgfβ1fl allele.

Primers were ordered from IDT in IDTE buffer and diluted to 10 μM in water before use. PCR was performed in a Multigene Optimax (Labnet) with the following conditions: 94°C for 5 mins, 35 cycles of 94°C for 30 sec, 53°C for 30 sec and 72°C for 30 sec, and finally 72°C for 10 mins. Products were separated by gel electrophoresis on a 1.5% agarose gel using 90 V for 80 minutes. The genetic sex of mice was determined using a previously described primer set and protocol (McFarlane et al., 2013).

Epidermal separation, extraction of RNA and generation of cDNA

Skins were excised from newborn pups and floated overnight at 4°C on 0.25% trypsin (Corning) containing ProtectRNA RNA Inhibitor (Sigma) at the manufacturer’s recommended dilution. The epidermis was peeled from the dermis and stored in RNAlater (Ambion). Precautions were taken against RNA degradation including treatment of all surfaces with RNase Erase (MP Bio). On day of harvest the epidermis was homogenized using autoclaved 1.5 mL tube pestles followed by centrifugation through a QiaShredder column (Qiagen). Total RNA was then extracted from the supernatant using an RNeasy Midi Plus kit (Qiagen) following the manufacturer’s protocol. The concentration and 260/280 values were determined by a Nanodrop 2000 (Thermo Scientific) and 2.5 μg of high purity (260/280 > 2.0) total RNA was reverse transcribed to cDNA using 200 units of M-MLV Reverse Transcriptase and random primers (Promega) following the manufacturer’s protocol.

Quantitative PCR

qPCR was performed using cDNA with a final concentration of 1/100 of the reverse transcription product. Genomic DNA was diluted 1/4. Primers were used at a final concentration of 250 nM along with PerfeCTa SYBR Green SuperMix for iQ (Quanta Biosciences) using the manufacturer’s recommended dilution. Reactions were run in a 20 μL reaction volume in 96 well format in a MyIQ cycler (BioRad) with the following conditions: 95°C for 10 mins, 45 cycles of 95°C for 10 sec and 55°C for 60 sec, followed by a melt curve.

List of qPCR primers (intron spanning)

Tgfβ1 - 5’ CACCGGAGAGCCCTGGATA, 5’ CCAAATATAGGGGCAGGGTCC

ΔNp63 – 5’ GAAAACAATGCCCAGACTCAA, 5’ TGCGCGTGGTCTGTGTTA

TAp63 - 5’ ACCCTTACATCCAGCGTTTCA, 5’ GCTGAGGAACTCGCTTGTCT

IL16 - 5’ GCCATTCAGCCTACACCAGT, 5’ GGAGGAGACTCAGGCAAACC

Ceacam2 - 5’ GCGAGGGGAAGAGGCATTT, 5’ AAAGACTGTCCTACCAGCGAG

Yif1b - 5’ AGCTAAACTTTGTTTCTCAGCCTA, 5’ CGTTTACACAGACGCAGCAC

CD22 - 5’ ATCCCCAAACCCTCTTTGCC, 5’ ATCCATCGTTCAGTCCCTGC

Gapdh – 5’ AAATGGTGAAGGTCGGTGTGAACG, 5’ TGGCAACAATCTCCACTTTGCCAC

Histology

Tissue was embedded in Optimal Cutting Temperature (OCT) compound (Sakura Tissue-Tek) and snap-frozen on 2-methylbutane and dry-ice. Skin sections were folded before embedding using a method adapted from (Paus et al., 1999). Tissue was cut at 10 μm on a CM1950 cryostat (Leica), adhered to Superfrost positively charged slides (Thermo Scientific) and frozen at −80°C.

Alternatively, tissue was fixed overnight in 10% Neutral Buffered Formalin (VWR) followed by 70% ethanol immersion for at least 12 hours, paraffin embedded in a TP1020 paraffin processor, cut in 6 μm sections and cured at 37°C overnight. Tissue sections were stained with hematoxylin and eosin (H&E) using an ST5010 Autostainer XL (Leica).

Hair was plucked from euthanized mice with forceps and placed on glass slides (VWR). Vectamount AQ mounting medium (Vector Labs) was added and the slides were coverslipped and sealed with nail varnish.

Immunofluorescence / immunohistochemistry

Frozen sections were thawed, fixed in 4% paraformaldehyde for 10 minutes and permeabilized with 0.1% Triton X100 in PBS for 10 minutes. Paraffin-embedded sections were deparaffinized in xylene washes, rehydrated in a ethanol gradient and washed in PBS. Slides were immersed in Antigen Unmasking Solution (Vector Labs) and steamed for 15 minutes. If blocking of endogenous peroxidase activity was required, slides were incubated in a 3% H2O2 solution for 15 minutes. Sections were blocked in 5% goat serum for 20 minutes and incubated with primary antibody overnight. Biotinylated secondary antibody was added the next morning, with anti-K14 if required. For immunofluorescence analysis, the appropriate streptavidin-fluorescent conjugate was added, slides were mounted with Vectashield mounting medium containing DAPI and sealed with nail varnish. Images were captured on a CKX41 microscope using cellSens software (Olympus). For immunohistochemistry, streptavidin-horse radish peroxidase (Vector Labs) was added and peroxidase was detected using diaminobenzidine (Vector Labs). Sections were counterstained with hematoxylin, dehydrated and mounted with xylene-substitute mounting medium (Shandon). Images were taken on a BX43 microscope using cellSens software (Olympus). All antibody/conjugate incubations were performed in 3% BSA in tris buffered saline with tween.

TUNEL assay

An In Situ Cell Death Detection Kit, Fluorescein (Roche) was used to detect apoptotic cells in frozen newborn skin sections according to the manufacturer’s recommended protocol.

Western blotting

Skins were excised from newborn pups and floated overnight at 4°C on 0.25% trypsin (Corning). The epidermis was peeled from the dermis, lysed, separated by SDS PAGE and immunoblotted as previously described (Blazanin et al., 2017).

Antibodies and fluorescent conjugates

Antigen Species Manufacturer Dilution Catalogue No.
α-Actin Mouse Cell Signaling Technologies 1:1000 3700
α-BLIMP1 Rat Santa Cruz 1:500 SC-47732
α-cKIT (D13A2 XP) Rabbit Cell Signaling Technologies 1:1000 3074
α-E-Cadherin (24E10) Rabbit Cell Signaling Technologies 1:1000 3195
α-Integrin a6 Rabbit Novus 1:500 NBP1-85747
α-Keratin 1 Rabbit Biolegend 1:500 905203
α-Keratin 5 Chicken Biolegend 1:500 905903
α-Keratin 10 Rabbit Biolegend 1:500 905403
α-Keratin 14 Chicken Biolegend 1:1000 906004
α-Ki67 (D3B5) Rabbit Cell Signaling Technologies 1:500 12202
α-LEF1 Rabbit Cell Signaling Technologies 1:1000 2230
α-Loricrin Rabbit Biolegend 1:500 905103
α-p63 (Internal, N2C1) Rabbit GeneTex 1:1000 GTX102425
α-ΔNp63 Rat Biolegend 1:1000 699501
α-TAp63 Mouse Biolegend 1:1000 938101
α-phospho- SMAD2 (S465/467) / SMAD3 (S423/425) D6G10 XP Rabbit Cell Signaling Technologies 1:1000 9510
α-SMAD2/3 Rabbit Millipore 1:1000 07-408
α-TGFβ1 Rabbit Abcam 1:20 ab215715
α-chicken-Alexa Fluor 568 Goat Invitrogen 1:500 A11041
α-rabbit (Biotinylated) Goat Vector Laboratories 1:500 BA-1000
α-rat (Biotinylated) Goat Vector Laboratories 1:500 BA-9401
Streptavidin-Alexa Fluor 488 N/A Invitrogen 1:500 S32354
Streptavidin-Alexa Fluor 568 N/A Invitrogen 1:500 S11226
Streptavidin-HRP N/A Vector Laboratories 1:1000 SA-5004
Open in a new tab

Image analysis

Linear measurements were taken on cellSens software (Olympus) and fluorescence-positive cells were quantified using QuPath 0.2.3 (Bankhead et al., 2017). The brush tool was used to trace continuous sections of cells along the basal layer in each of the images. Once desired sections were traced, the positive cell detection analysis feature was used to give the total number of cells and number of positive cells within the traced region. Default settings were used on all parameters with the exception of the intensity threshold, cell expansion, and score compartment.

Statistical analysis

Data was managed in Excel (Microsoft), and analyzed and presented in Prism 8 (Graphpad). Student’s T Test was performed using unpaired, two-tailed analysis with assumed Gaussian distribution and considered significant if p < 0.05. * = p < 0.05; ** = p < 0.01; *** = p < 0.001.

  • Keratinocyte-derived TGFβ1 is required for the development of hair follicles and epidermis.

  • Hair cycling and epidermal homeostasis were also disrupted in TGFβ1 knockout mice.

  • A new role for TGFβ1 in melanocyte migration was also revealed.

Acknowledgements

This work was supported by the USDA National Institute of Food and Agriculture Federal Appropriations under Project PEN04607 and Accession number 1009993, and funding from the Department of Veterinary and Biomedical Sciences, Pennsylvania State University to AG. Initial work was supported by NIH grant R01 CA122109 and the Elsa U. Pardee Foundation to AG. FC was supported by a Postdoctoral Fellowship award from the Penn State Cancer Institute. The authors wish to acknowledge Dr. Stuart Yuspa for continued support.

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Competing interests

The authors declare no competing interests.

Bibliography

  1. Azhar M, Yin M, Bommireddy R, Duffy JJ, Yang J, Pawlowski SA, Boivin GP, Engle SJ, Sanford LP, Grisham C, Singh RR, Babcock GF and Doetschman T (2009) ‘Generation of mice with a conditional allele for transforming growth factor beta 1 gene’, Genesis, 47(6), pp. 423–431. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Bankhead P, Loughrey MB, Fernandez JA, Dombrowski Y, McArt DG, Dunne PD, McQuaid S, Gray RT, Murray LJ, Coleman HG, James JA, Salto-Tellez M and Hamilton PW (2017) ‘QuPath: Open source software for digital pathology image analysis’, Sci Rep, 7(1), pp. 16878. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Barbieri CE, Tang LJ, Brown KA and Pietenpol JA (2006) ‘Loss of p63 leads to increased cell migration and up-regulation of genes involved in invasion and metastasis’, Cancer Res, 66(15), pp. 7589–97. [DOI] [PubMed] [Google Scholar]
  4. Blazanin N, Son J, Craig-Lucas AB, John CL, Breech KJ, Podolsky MA and Glick AB (2017) ‘ER stress and distinct outputs of the IRE1α RNase control proliferation and senescence in response to oncogenic Ras’, Proc Natl Acad Sci U S A, 114(37), pp. 9900–9905. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Botchkarev VA, Botchkareva NV, Roth W, Nakamura M, Chen LH, Herzog W, Lindner G, McMahon JA, Peters C, Lauster R, McMahon AP and Paus R (1999) ‘Noggin is a mesenchymally derived stimulator of hair-follicle induction’, Nat Cell Biol, 1(3), pp. 158–64. [DOI] [PubMed] [Google Scholar]
  6. Botchkareva NV, Khlgatian M, Longley BJ, Botchkarev VA and Gilchrest BA (2001) ‘SCF/c-kit signaling is required for cyclic regeneration of the hair pigmentation unit’, FASEB J, 15(3), pp. 645–58. [DOI] [PubMed] [Google Scholar]
  7. Bui NHB, Napoli M, Davis AJ, Abbas HA, Rajapakshe K, Coarfa C and Flores ER (2020) ‘Spatiotemporal Regulation of ΔNp63 by TGFβ-Regulated miRNAs Is Essential for Cancer Metastasis’, Cancer Res, 80(13), pp. 2833–2847. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Cordray P and Satterwhite DJ (2005) ‘TGF-beta induces novel Lef-1 splice variants through a Smad-independent signaling pathway’, Dev Dyn, 232(4), pp. 969–78. [DOI] [PubMed] [Google Scholar]
  9. Dassule HR, Lewis P, Bei M, Maas R and McMahon AP (2000) ‘Sonic hedgehog regulates growth and morphogenesis of the tooth’, Development, 127(22), pp. 4775–85. [DOI] [PubMed] [Google Scholar]
  10. Ding X, Park SI, McCauley LK and Wang CY (2013) ‘Signaling between transforming growth factor β (TGF-β) and transcription factor SNAI2 represses expression of microRNA miR-203 to promote epithelial-mesenchymal transition and tumor metastasis’, J Biol Chem, 288(15), pp. 10241–53. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Dubois CM, Ruscetti FW, Stankova J and Keller JR (1994) ‘Transforming growth factor-beta regulates c-kit message stability and cell-surface protein expression in hematopoietic progenitors’, Blood, 83(11), pp. 3138–45. [PubMed] [Google Scholar]
  12. Foitzik K, Lindner G, Mueller-Roever S, Maurer M, Botchkareva N, Botchkarev V, Handjiski B, Metz M, Hibino T, Soma T, Dotto GP and Paus R (2000) ‘Control of murine hair follicle regression (catagen) by TGF-beta1 in vivo’, FASEB J, 14(5), pp. 752–60. [DOI] [PubMed] [Google Scholar]
  13. Foitzik K, Paus R, Doetschman T and Dotto GP (1999) ‘The TGF-beta2 isoform is both a required and sufficient inducer of murine hair follicle morphogenesis’, Dev Biol, 212(2), pp. 278–89. [DOI] [PubMed] [Google Scholar]
  14. Fukunishi N, Katoh I, Tomimori Y, Tsukinoki K, Hata R, Nakao A, Ikawa Y and Kurata S (2010) ‘Induction of ΔNp63 by the newly identified keratinocyte-specific transforming growth factor β Signaling Pathway with Smad2 and IκB Kinase α in squamous cell carcinoma’, Neoplasia, 12(12), pp. 969–79. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Glick AB, Kulkarni AB, Tennenbaum T, Hennings H, Flanders KC, O’Reilly M, Sporn MB, Karlsson S and Yuspa SH (1993) ‘Loss of expression of transforming growth factor beta in skin and skin tumors is associated with hyperproliferation and a high risk for malignant conversion’, Proc Natl Acad Sci U S A, 90(13), pp. 6076–80. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Guasch G, Schober M, Pasolli HA, Conn EB, Polak L and Fuchs E (2007) ‘Loss of TGFbeta signaling destabilizes homeostasis and promotes squamous cell carcinomas in stratified epithelia’, Cancer Cell, 12(4), pp. 313–327. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Hansen LA, Alexander N, Hogan ME, Sundberg JP, Dlugosz A, Threadgill DW, Magnuson T and Yuspa SH (1997) ‘Genetically null mice reveal a central role for epidermal growth factor receptor in the differentiation of the hair follicle and normal hair development’, Am J Pathol, 150(6), pp. 1959–75. [PMC free article] [PubMed] [Google Scholar]
  18. Hardman MJ, Sisi P, Banbury DN and Byrne C (1998) ‘Patterned acquisition of skin barrier function during development’, Development, 125(8), pp. 1541–52. [DOI] [PubMed] [Google Scholar]
  19. Inoue K, Aoi N, Yamauchi Y, Sato T, Suga H, Eto H, Kato H, Tabata Y and Yoshimura K (2009) ‘TGF-beta is specifically expressed in human dermal papilla cells and modulates hair folliculogenesis’, J Cell Mol Med, 13(11–12), pp. 4643–56. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Kim DS, Park SH and Park KC (2004) ‘Transforming growth factor-beta1 decreases melanin synthesis via delayed extracellular signal-regulated kinase activation’, Int J Biochem Cell Biol, 36(8), pp. 1482–91. [DOI] [PubMed] [Google Scholar]
  21. Koster MI, Dai D, Marinari B, Sano Y, Costanzo A, Karin M and Roop DR (2007) ‘p63 induces key target genes required for epidermal morphogenesis’, Proc Natl Acad Sci U S A, 104(9), pp. 3255–60. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Koster MI, Kim S, Mills AA, DeMayo FJ and Roop DR (2004) ‘p63 is the molecular switch for initiation of an epithelial stratification program’, Genes Dev, 18(2), pp. 126–31. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Kulkarni AB, Huh CG, Becker D, Geiser A, Lyght M, Flanders KC, Roberts AB, Sporn MB, Ward JM and Karlsson S (1993) ‘Transforming growth factor beta 1 null mutation in mice causes excessive inflammatory response and early death’, Proc Natl Acad Sci U S A, 90(2), pp. 770–4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Kunisada T, Yoshida H, Yamazaki H, Miyamoto A, Hemmi H, Nishimura E, Shultz LD, Nishikawa S and Hayashi S (1998) ‘Transgene expression of steel factor in the basal layer of epidermis promotes survival, proliferation, differentiation and migration of melanocyte precursors’, Development, 125(15), pp. 2915–23. [DOI] [PubMed] [Google Scholar]
  25. Lena AM, Shalom-Feuerstein R, Rivetti di Val Cervo P, Aberdam D, Knight RA, Melino G and Candi E (2008) ‘miR-203 represses ‘stemness’ by repressing DeltaNp63’, Cell Death Differ, 15(7), pp. 1187–95. [DOI] [PubMed] [Google Scholar]
  26. Lesko MH, Driskell RR, Kretzschmar K, Goldie SJ and Watt FM (2013) ‘Sox2 modulates the function of two distinct cell lineages in mouse skin’, Dev Biol, 382(1), pp. 15–26. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Lichtenberger BM, Mastrogiannaki M and Watt FM (2016) ‘Epidermal β-catenin activation remodels the dermis via paracrine signalling to distinct fibroblast lineages’, Nat Commun, 7, pp. 10537. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Liu X, Alexander V, Vijayachandra K, Bhogte E, Diamond I and Glick A (2001) ‘Conditional epidermal expression of TGFbeta 1 blocks neonatal lethality but causes a reversible hyperplasia and alopecia’, Proc Natl Acad Sci U S A, 98(16), pp. 9139–44. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Lyons RM and Moses HL (1990) ‘Transforming growth factors and the regulation of cell proliferation’, Eur J Biochem, 187(3), pp. 467–73. [DOI] [PubMed] [Google Scholar]
  30. Marie JC, Liggitt D and Rudensky AY (2006) ‘Cellular mechanisms of fatal early-onset autoimmunity in mice with the T cell-specific targeting of transforming growth factor-beta receptor’, Immunity, 25(3), pp. 441–54. [DOI] [PubMed] [Google Scholar]
  31. Massagué J (1998) ‘TGF-beta signal transduction’, Annu Rev Biochem, 67, pp. 753–91. [DOI] [PubMed] [Google Scholar]
  32. McFarlane L, Truong V, Palmer JS and Wilhelm D (2013) ‘Novel PCR assay for determining the genetic sex of mice’, Sex Dev, 7(4), pp. 207–11. [DOI] [PubMed] [Google Scholar]
  33. McKeown ST, Hyland PL, Locke M, Mackenzie IC and Irwin CR (2003) ‘Keratinocyte growth factor and scatter factor expression by regionally defined oral fibroblasts’, Eur J Oral Sci, 111(1), pp. 42–50. [DOI] [PubMed] [Google Scholar]
  34. McNairn AJ, Doucet Y, Demaude J, Brusadelli M, Gordon CB, Uribe-Rivera A, Lambert PF, Bouez C, Breton L and Guasch G (2013) ‘TGFβ signaling regulates lipogenesis in human sebaceous glands cells’, BMC Dermatol, 13, pp. 2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Mills AA, Zheng B, Wang XJ, Vogel H, Roop DR and Bradley A (1999) ‘p63 is a p53 homologue required for limb and epidermal morphogenesis’, Nature, 398(6729), pp. 708–713. [DOI] [PubMed] [Google Scholar]
  36. Murakami M, Matsuzaki F and Funaba M (2009) ‘Regulation of melanin synthesis by the TGF-beta family in B16 melanoma cells’, Mol Biol Rep, 36(6), pp. 1247–50. [DOI] [PubMed] [Google Scholar]
  37. Naruse T, Aoki M, Fujimoto N, Arase S, Oura H, Ueda Y and Ikeda A (2017) ‘Novel ALK5 inhibitor TP0427736 reduces TGF-β induced growth inhibition in human outer root sheath cells and elongates anagen phase in mouse hair follicles’, Pharmacol Rep, 69(3), pp. 485–491. [DOI] [PubMed] [Google Scholar]
  38. Nishimura EK, Suzuki M, Igras V, Du J, Lonning S, Miyachi Y, Roes J, Beermann F and Fisher DE (2010) ‘Key roles for transforming growth factor beta in melanocyte stem cell maintenance’, Cell Stem Cell, 6(2), pp. 130–40. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Ogawa K, Saito A, Matsui H, Suzuki H, Ohtsuka S, Shimosato D, Morishita Y, Watabe T, Niwa H and Miyazono K (2007) ‘Activin-Nodal signaling is involved in propagation of mouse embryonic stem cells’, J Cell Sci, 120(Pt 1), pp. 55–65. [DOI] [PubMed] [Google Scholar]
  40. Oshimori N and Fuchs E (2012) ‘Paracrine TGF-β signaling counterbalances BMP-mediated repression in hair follicle stem cell activation’, Cell Stem Cell, 10(1), pp. 63–75. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Parsa R, Yang A, McKeon F and Green H (1999) ‘Association of p63 with proliferative potential in normal and neoplastic human keratinocytes’, J Invest Dermatol, 113(6), pp. 1099–105. [DOI] [PubMed] [Google Scholar]
  42. Paus R, Foitzik K, Welker P, Bulfone-Paus S and Eichmüller S (1997) ‘Transforming growth factor-beta receptor type I and type II expression during murine hair follicle development and cycling’, J Invest Dermatol, 109(4), pp. 518–26. [DOI] [PubMed] [Google Scholar]
  43. Paus R, Müller-Röver S, Van Der Veen C, Maurer M, Eichmüller S, Ling G, Hofmann U, Foitzik K, Mecklenburg L and Handjiski B (1999) ‘A comprehensive guide for the recognition and classification of distinct stages of hair follicle morphogenesis’, J Invest Dermatol, 113(4), pp. 523–32. [DOI] [PubMed] [Google Scholar]
  44. Paus R, van der Veen C, Eichmüller S, Kopp T, Hagen E, Müller-Röver S and Hofmann U (1998) ‘Generation and cyclic remodeling of the hair follicle immune system in mice’, J Invest Dermatol, 111(1), pp. 7–18. [DOI] [PubMed] [Google Scholar]
  45. Peters EM, Tobin DJ, Botchkareva N, Maurer M and Paus R (2002) ‘Migration of melanoblasts into the developing murine hair follicle is accompanied by transient c-Kit expression’, J Histochem Cytochem, 50(6), pp. 751–66. [DOI] [PubMed] [Google Scholar]
  46. Phan QM, Fine GM, Salz L, Herrera GG, Wildman B, Driskell IM and Driskell RR (2020) ‘Lef1 expression in fibroblasts maintains developmental potential in adult skin to regenerate wounds’, Elife, 9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Reichelt J, Büssow H, Grund C and Magin TM (2001) ‘Formation of a normal epidermis supported by increased stability of keratins 5 and 14 in keratin 10 null mice’, Mol Biol Cell, 12(6), pp. 1557–68. [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Sellheyer K, Bickenbach JR, Rothnagel JA, Bundman D, Longley MA, Krieg T, Roche NS, Roberts AB and Roop DR (1993) ‘Inhibition of skin development by overexpression of transforming growth factor beta 1 in the epidermis of transgenic mice’, Proc Natl Acad Sci U S A, 90(11), pp. 5237–41. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Senoo M, Pinto F, Crum CP and McKeon F (2007) ‘p63 Is essential for the proliferative potential of stem cells in stratified epithelia’, Cell, 129(3), pp. 523–36. [DOI] [PubMed] [Google Scholar]
  50. Shull MM, Ormsby I, Kier AB, Pawlowski S, Diebold RJ, Yin M, Allen R, Sidman C, Proetzel G and Calvin D (1992) ‘Targeted disruption of the mouse transforming growth factor-beta 1 gene results in multifocal inflammatory disease’, Nature, 359(6397), pp. 693–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Sonnenberg A, Calafat J, Janssen H, Daams H, van der Raaij-Helmer LM, Falcioni R, Kennel SJ, Aplin JD, Baker J and Loizidou M (1991) ‘Integrin alpha 6/beta 4 complex is located in hemidesmosomes, suggesting a major role in epidermal cell-basement membrane adhesion’, J Cell Biol, 113(4), pp. 907–17. [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. Sundqvist A, Vasilaki E, Voytyuk O, Bai Y, Morikawa M, Moustakas A, Miyazono K, Heldin CH, Ten Dijke P and van Dam H (2020) ‘TGFβ and EGF signaling orchestrates the AP-1- and p63 transcriptional regulation of breast cancer invasiveness’, Oncogene, 39(22), pp. 4436–4449. [DOI] [PMC free article] [PubMed] [Google Scholar]
  53. Takayama H, La Rochelle WJ, Anver M, Bockman DE and Merlino G (1996) ‘Scatter factor/hepatocyte growth factor as a regulator of skeletal muscle and neural crest development’, Proc Natl Acad Sci U S A, 93(12), pp. 5866–71. [DOI] [PMC free article] [PubMed] [Google Scholar]
  54. Telerman SB, Rognoni E, Sequeira I, Pisco AO, Lichtenberger BM, Culley OJ, Viswanathan P, Driskell RR and Watt FM (2017) ‘Dermal Blimp1 Acts Downstream of Epidermal TGFβ and Wnt/β-Catenin to Regulate Hair Follicle Formation and Growth’, J Invest Dermatol, 137(11), pp. 2270–2281. [DOI] [PMC free article] [PubMed] [Google Scholar]
  55. Thatcher EJ, Paydar I, Anderson KK and Patton JG (2008) ‘Regulation of zebrafish fin regeneration by microRNAs’, Proc Natl Acad Sci U S A, 105(47), pp. 18384–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  56. Titen SWA, Johnson MTB, Capecchi M and Golic KG (2020) ‘Site-Specific Recombination with Inverted Target Sites: A Cautionary Tale of Dicentric and Acentric Chromosomes’, Genetics, 215(4), pp. 923–930. [DOI] [PMC free article] [PubMed] [Google Scholar]
  57. Truong AB, Kretz M, Ridky TW, Kimmel R and Khavari PA (2006) ‘p63 regulates proliferation and differentiation of developmentally mature keratinocytes’, Genes Dev, 20(22), pp. 3185–97. [DOI] [PMC free article] [PubMed] [Google Scholar]
  58. van Genderen C, Okamura RM, Fariñas I, Quo RG, Parslow TG, Bruhn L and Grosschedl R (1994) ‘Development of several organs that require inductive epithelial-mesenchymal interactions is impaired in LEF-1-deficient mice’, Genes Dev, 8(22), pp. 2691–703. [DOI] [PubMed] [Google Scholar]
  59. Velegraki M, Salem M, Ansa-Addo EA, Wu BX and Li Z (2021) ‘Autocrine transforming growth factor β1 in regulatory T cell biology-gone but not missed’, Immunity, 54(3), pp. 395–396. [DOI] [PubMed] [Google Scholar]
  60. Wang Y, Du J, Niu X, Fu N, Wang R, Zhang Y, Zhao S, Sun D and Nan Y (2017) ‘MiR-130a-3p attenuates activation and induces apoptosis of hepatic stellate cells in nonalcoholic fibrosing steatohepatitis by directly targeting TGFBR1 and TGFBR2’, Cell Death Dis, 8(5), pp. e2792. [DOI] [PMC free article] [PubMed] [Google Scholar]
  61. Yang A, Kaghad M, Wang Y, Gillett E, Fleming MD, Dötsch V, Andrews NC, Caput D and McKeon F (1998) ‘p63, a p53 homolog at 3q27–29, encodes multiple products with transactivating, death-inducing, and dominant-negative activities’, Mol Cell, 2(3), pp. 305–16. [DOI] [PubMed] [Google Scholar]
  62. Yang A, Schweitzer R, Sun D, Kaghad M, Walker N, Bronson RT, Tabin C, Sharpe A, Caput D, Crum C and McKeon F (1999) ‘p63 is essential for regenerative proliferation in limb, craniofacial and epithelial development’, Nature, 398(6729), pp. 714–8. [DOI] [PubMed] [Google Scholar]
  63. Yang G, Li Y, Nishimura EK, Xin H, Zhou A, Guo Y, Dong L, Denning MF, Nickoloff BJ and Cui R (2008) ‘Inhibition of PAX3 by TGF-beta modulates melanocyte viability’, Mol Cell, 32(4), pp. 554–63. [DOI] [PubMed] [Google Scholar]
  64. Yang Y, Zenke Y, Hirai T and Kaplan DH (2019) ‘Keratinocyte-derived TGFβ is not required to maintain skin immune homeostasis’, J Dermatol Sci, 94(2), pp. 290–297. [DOI] [PMC free article] [PubMed] [Google Scholar]
  65. Yi X, Li X, Zhou Y, Ren S, Wan W, Feng G and Jiang X (2014) ‘Hepatocyte growth factor regulates the TGF-β1-induced proliferation, differentiation and secretory function of cardiac fibroblasts’, Int J Mol Med, 34(2), pp. 381–90. [DOI] [PMC free article] [PubMed] [Google Scholar]
  66. Yoshimura A, Wakabayashi Y and Mori T (2010) ‘Cellular and molecular basis for the regulation of inflammation by TGF-beta’, J Biochem, 147(6), pp. 781–92. [DOI] [PMC free article] [PubMed] [Google Scholar]
  67. Zhou P, Byrne C, Jacobs J and Fuchs E (1995) ‘Lymphoid enhancer factor 1 directs hair follicle patterning and epithelial cell fate’, Genes Dev, 9(6), pp. 700–13. [DOI] [PubMed] [Google Scholar]

RESOURCES