Aplasia cutis congenita pathomechanisms reveal key regulators of skin and skin appendage morphogenesis

Abstract

Aplasia cutis congenita (ACC) manifests at birth as a defect of the scalp skin. New findings answer two longstanding questions: why ACC forms and why it affects mainly the midline scalp skin. Dominant-negative mutations in the genes KCTD1 or KCTD15 cause ACC due to loss of function of KCTD1/KCTD15 complexes in cranial neural crest cells (NCCs), which normally form midline cranial suture mesenchymal cells that express keratinocyte growth factors. Loss of KCTD1/KCTD15 function in NCCs impairs the formation of normal midline cranial sutures and, consequently, the overlying skin, resulting in ACC. Moreover, KCTD1/KCTD15 complexes in keratinocytes regulate skin appendage morphogenesis.

INTRODUCTION

The mechanisms that regulate scalp skin morphogenesis during development are not well understood. Cranial bones and cartilage structures are derived either from mesodermal cells or from cranial neural crest cells. Studies in mice have shown that cranial neural crest cells (NCCs) give rise to midline cranial sutures (interfrontal and sagittal), as well as several midline cranial bones (including nasal, frontal, and interparietal bones), mandible and maxilla, and several other cranial bones. In contrast, the coronal suture, parietal bones, occipital bones, and several other bones are of mesodermal origin (Mishina and Snider, 2014). The epidermis is of neuroectodermal origin and develops from a single layer of multipotent epithelial cells during embryogenesis. Mesenchymal cells of the dermis are derived from the dermomyotome and interact with the overlying epithelium to form hair placodes (Fuchs, 2007). Whether inductive signaling mechanisms from neural crest-derived cranial structures to the overlying epithelial cells of the forming epidermis orchestrate proper scalp skin formation during development is an important open question. In this context, identifying the pathomechanisms that cause aplasia cutis congenita (ACC), a congenital scalp skin defect, may reveal important insights into the cellular and molecular mechanisms that are required for proper scalp skin development and may uncover potential inductive signaling mechanisms between cell populations of different embryonic origin that are required for scalp skin formation.

ACC manifests at birth as a localized scalp skin defect and affects mostly the scalp vertex and the skin along the midline cranial sutures (Bessis et al., 2017). There can be either a thinning of the epidermis with loss of skin appendages (membranous ACC), or a full-thickness skin defect. ACC can be associated with underlying skull defects, a hair collar sign, and meningeal heterotopia (Bessis et al., 2017). Previous clinical classifications have used the term ACC not only for congenital scalp skin defects but also for localized skin defects at other anatomic sites found at birth (Frieden, 1986). Here, ACC is considered to be a disease manifestation that affects only the scalp skin, whereas congenital skin defects at other anatomic sites are viewed as distinct clinical entities with a different pathogenesis that should be described with other terminology. The rationale for this distinction is provided by new data from Raymundo et al (2023), which demonstrate that mutations in the transcriptional regulators KCTD1 or KCTD15 lead to scalp ACC as a secondary consequence of defects in cranial NCCs that form mesenchymal cells of midline cranial sutures which normally produce several keratinocyte growth factors. These data link ACC formation in patients with KCTD1 or KCTD15 mutations spatiotemporally to abnormalities in the formation of NCC-derived midline cranial sutures during development, establishing ACC as a neurocristopathy and demonstrating that ACC is not a consequence of a primary keratinocyte abnormality in these patients. Thus, it is proposed here that the term ACC be used only for congenital skin defects that are limited to the scalp skin.

ACC most commonly occurs as a sporadic manifestation without other congenital abnormalities. Infrequently, it can occur with a familial inheritance pattern, as already described centuries ago (Campbell, 1826). A gene mutation in the ribosomal GTPase BMS1 was identified in a family with autosomal-dominant ACC that did not have other clinical abnormalities (Marneros, 2013).

The cellular origin and the pathomechanisms that may explain why ACC affects mainly the midline scalp have been elusive. Previous hypotheses suggested that ACC occurs on the scalp due to a primary keratinocyte abnormality that may impair keratinocyte migration or proliferation at a time when the embryonic head rapidly expands and when the demand on keratinocytes is particularly high (Stephan et al., 1982; Sutton, 1935). Notably, ACC can occur as part of various rare genetic syndromes (Marneros, 2015). For example, Adams-Oliver syndrome manifests mainly with ACC and terminal transverse limb defects and has been linked to mutations in genes that affect Notch signaling (EOGT, NOTCH1, RBPJ, DLL4) (Hassed et al., 2012; Meester et al., 2015; Shaheen et al., 2013; Stittrich et al., 2014) or Cdc42/rac1 activities (Shaheen et al., 2011; Southgate et al., 2011). While the identification of these gene mutations implicates these genes and pathways in proper scalp skin morphogenesis, it does not identify the cell types whose defects lead to ACC.

Mutations in KCTD1 were identified in individuals with Scalp–Ear–Nipple (SEN) syndrome (also known as Finlay-Marks syndrome), which manifests primarily as scalp ACC (sometimes associated with an underlying skull bone defect), abnormalities of the ears, and absence or hypoplasia of nipples and breasts (Edwards et al., 1994; Finlay and Marks, 1978; Marneros et al., 2013). Additional reported abnormalities include sparse hair, anhidrosis, nose abnormalities, dental abnormalities, and kidney disease (Marneros, 2020; Marneros et al., 2013; Wang et al., 2023). Combining human genetics with research using genetic mouse models has now revealed the cellular origin and the pathomechanisms that lead to ACC in patients with SEN syndrome (Raymundo et al., 2023), which will be discussed here. These findings likely share common cellular and molecular pathomechanisms with cases of ACC that are not part of SEN syndrome.

ACC IN SEN SYNDROME IS DUE TO HETEROZYGOUS MISSENSE MUTATIONS IN KCTD1

Heterozygous missense mutations in KCTD1 were found with an autosomal-dominant inheritance pattern in families with SEN syndrome (Marneros et al., 2013). KCTD1 belongs to the family of bric-à-brac, tramtrack, broad complex (BTB) domain–containing proteins (Ji et al., 2016). Little is known about the cellular functions of KCTD1. It represses the transactivation of the transcription factor AP-2α through binding via its BTB domain, and KCTD1 mutations in SEN syndrome, mainly within the BTB domain, abrogate the binding and inhibitory activity of KCTD1 on AP-2α and result in increased transcriptional activity of AP-2α (Ding et al., 2009; Hu et al., 2020; Smaldone et al., 2019; Raymundo et al., 2023). Moreover, KCTD1 also inhibits the activity of AP-2β and AP-2γ in vitro (Ding et al., 2009). Notably, a highly homologous paralogue of KCTD1, KCTD15, has also been shown to inhibit AP-2 activity (Zarelli et al., 2013). Various studies also implicated KCTD1 and KCTD15 as regulators of other signaling pathways or molecular targets. For example, it has been suggested that KCTD1 and KCTD15 affect Shh signaling (Di Fiore et al., 2023; Spiombi et al., 2019) or that they regulate Wnt/β-catenin activity (Dutta and Dawid, 2010; Hu et al., 2020). To what extent KCTD1 and/or KCTD15 exert specific cellular function through these distinct signaling mechanisms in specific cell types in vivo remains to be determined.

To assess the in vivo roles of KCTD1, Kctd1 null mice were generated and carefully examined (Marneros, 2020). Inactivation of Kctd1 in mice did not phenocopy the skin abnormalities that were observed in SEN syndrome patients (Marneros, 2020). However, Kctd1 null mice or mice lacking Kctd1 specifically in the kidney epithelium (Six2Cre+Kctd1fl/fl mice) developed abnormalities of the distal nephron that resulted in an impaired terminal differentiation of the distal convoluted tubules (DCTs), which play important roles in urinary salt reabsorption. This caused a severe salt-losing tubulopathy and secondary bone abnormalities due to aberrant calcium and magnesium transport processes in the distal nephron (Marneros, 2020, 2021). The terminal differentiation defect of DCTs was associated with hyperactivation of β-catenin/mTOR signaling in DCT epithelial cells that resulted in progressive renal fibrosis. The transcription factor AP-2β was found to be required for the formation of early-stage DCTs during nephrogenesis and to induce expression of Kctd1, which subsequently promoted the terminal differentiation of these early-stage DCTs (Lamontagne et al., 2022; Marneros, 2020). Inducible inactivation of AP-2β or Kctd1 in adult mice also resulted in these kidney abnormalities, albeit with a less severe manifestation. Notably, a subset of SEN syndrome patients also developed renal fibrosis with age (Marneros, 2020). Collectively, these findings in SEN syndrome patients with KCTD1 mutations and in mice lacking Kctd1 uncover a critical role of this gene for the proper development of the distal nephron and for the maintenance of its terminal differentiation state in adults.

KCTD1 CAN FORM HETEROMERIC COMPLEXES WITH KCTD15

The absence of a skin phenotype in Kctd1 null mice, while these mice display renal abnormalities that resemble those seen in SEN syndrome patients, suggests that a lack of Kctd1 can be compensated for in the skin but not in the distal nephron of the kidney. Raymundo et al. (2023) uncovered such a compensation mechanism: KCTD1 forms pentameric complexes with itself but also with its highly homologous paralogue KCTD15 (Figure 1a–c), and loss of KCTD1 can be compensated for by KCTD15 and vice versa depending on the abundance of KCTD1 and KCTD15 in a specific cell type (Ji et al., 2016; Raymundo et al., 2023). Modeling of the structure of KCTD15, based on the KCTD1 crystal structure, shows a very similar structure for both paralogues that allows KCTD1 and KCTD15 monomers to assemble into pentameric ring structures in any stochiometry, mediated through interactions among C-terminal domains and among BTB domains (Figure 1a–c). The KCTD1 mutations observed in patients with SEN syndrome would be predicted to destabilize this pentameric assembly by disrupting interactions between BTB domains (Smaldone et al., 2019).

Figure 1: KCTD1 and KCTD15 form pentameric ring structures as homomers or heteromers.

a. Crystal structure of KCTD1 reveals a pentameric organization of KCTD1 complexes. Interactions between monomers occur at the C-terminal domain (CTD) and the BTB domain.

b. AlphaFold modeling for KCTD15, based on the KCTD1 crystal structure, reveals a highly similar structural organization of KCTD15, which is a highly homologous paralogue of KCTD1.

c. Prediction of pentameric heteromers between KCTD1 and KCTD15 (in this example 3 KCTD1 units assemble with 2 KCTD15 units).

d. The p.Asp104His KCTD15 mutation disrupts an inter-subunit salt bridge with Arg118 of the adjacent KCTD15 or with the equivalent Arg92 of the KCTD1 subunit. Left: Asp104 (blue) at the inter-subunit interface of the KCTD15 pentamer (BTB domains). Middle: Asp104 forms an inter-subunit salt bridge with Arg118 between KCTD15 subunits. Right: similar electrostatic interaction formed by Asp104 (KCTD15) and Arg92 (KCTD1), which is the residue that is equivalent to Arg118 of KCTD15, observed at the KCTD1-KCTD15 interface of the hetero-pentamer.

Images from Raymundo et al (2023).

Link to license: https://creativecommons.org/licenses/by/4.0/deed.en

In DCT epithelial cells, KCTD1, but not KCTD15, is highly expressed; thus, loss of KCTD1 function cannot be compensated for by KCTD15 in DCTs, explaining why mice lacking Kctd1 in DCTs develop abnormalities in this epithelium. In contrast, in keratinocytes and cranial neural crest cells (NCCs) both KCTD1 and KCTD15 are expressed and can compensate for the other’s loss, explaining why Kctd1 null mice do not develop skin abnormalities.

The observation that SEN syndrome manifests in families in an autosomal-dominant pattern due to heterozygous missense mutations in KCTD1 suggests that these mutations have a dominant-negative effect that leads to a loss of function of KCTD1/KCTD15 complexes. Indeed, Raymundo et al. show that SEN syndrome KCTD1 mutations result in an increased propensity of KCTD1 mutant proteins to form amyloid-like aggregates that are unable to bind to AP-2α (a functional read-out of KCTD1 activity). These KCTD1 mutant proteins sequester and inactivate not only wild-type KCTD1 but also wild-type KCTD15, resulting in a combined inactivation of KCTD1 and KCTD15, consistent with a dominant-negative effect. Thus, in SEN syndrome, KCTD1 mutations would be predicted to result in a combined loss of KCTD1 and KCTD15 function in cells in which both are expressed, explaining why these patients develop skin abnormalities that are not observed in Kctd1 null mice.

Notably, a heterozygous mutation in KCTD15 that results in a KCTD15D104H mutant protein has been reported in a family that had ACC (associated with a midline bony skull defect); also present in the family were nose abnormalities and anosmia, dental anomalies, a cranial mass, and congenital cardiac defects (tetralogy of Fallot and patent ductus arteriosus; Miller, 2024; Miller, 2018). This p.Asp104His KCTD15 mutation disrupts an inter-subunit salt bridge with Arg118 of the adjacent KCTD15 or with the equivalent Arg92 of the KCTD1 subunit, which would be predicted to destabilize pentameric assembly (Figure 1d), similarly as reported for KCTD1 mutations found in SEN syndrome (Smaldone et al., 2019). Raymundo et al showed that this mutation impairs pentamer formation of KCTD15 proteins and leads to amyloid-like aggregate formation, as observed for the KCTD1 mutants. The KCTD15D104H mutant protein inactivated not only wild-type KCTD15 but also wild-type KCTD1. The observation that ACC occurs in patients with either dominant-negative KCTD1 or KCTD15 mutations and that mutant KCTD1 or KCTD15 proteins can both inactivate their wild-type paralogues is consistent with the hypothesis that these mutations result in a combined inactivation of KCTD1 and KCTD15 and that both wild-type proteins can compensate for the other’s loss.

KCTD1 AND KCTD15 ARE CRITICAL REGULATORS OF SKIN APPENDAGE MORPHOGENESIS

In addition to ACC, SEN syndrome patients often also have anhidrosis and sparse hair. Inactivation of Kctd1 or Kctd15 in murine keratinocytes, which express both to a similar extent, does not result in a phenotype. However, combined inactivation of both Kctd1 and Kctd15 in keratinocytes (K14Cre+Kctd1fl/flKctd15fl/fl mice) impaired proper hair morphogenesis and resulted in delayed postnatal hair growth with irregular hair morphology (e.g., curly whiskers) and sparse hair. Hair sparseness was also observed in older K14Cre+Kctd1fl/flKctd15fl/fl mice. Lack of Kctd1 and Kctd15 in keratinocytes also resulted in a postnatal growth retardation and in a delay of interdigital web space formation. Strikingly, these double mutant mice were found to also have severely diminished eccrine sweat glands and sebaceous glands (Raymundo et al., 2023). Sebaceous glands were smaller in size but still expressed terminal differentiation markers, indicating that the reduction in sebaceous glands was not due to a differentiation defect. The skin appendage abnormalities in K14Cre+Kctd1fl/flKctd15fl/fl mice are consistent with overlapping and compensatory functions of Kctd1 and Kctd15 in keratinocytes and explain why patients with dominant-negative KCTD1 or KCTD15 mutations develop hair abnormalities and anhidrosis. They also provide in vivo evidence that these mutations result in a combined loss-of-function of both KCTD1 and KCTD15.

Several SEN syndrome families have been reported and clinically evaluated, including some whose pedigrees span several generations (Edwards et al., 1994; Finlay and Marks, 1978; Marneros, 2020; Marneros et al., 2013; Wang et al., 2023). Thus, the spectrum of clinical abnormalities that result from dominant-negative KCTD1 missense mutations is rather well documented. In contrast, KCTD15 mutations have been reported only in a sporadic case and a small nuclear family with two affected members (Miller, 2024; Miller, 2018). Based on the in vitro and in vivo findings by Raymundo et al., it is likely that the identification of more individuals with KCTD15 mutations will reveal more overlap in skin and skin appendage defects between those patients and individuals with SEN syndrome. The findings in K14Cre+Kctd1fl/flKctd15fl/fl mice demonstrate that KCTD1/KCTD15 complexes are critical regulators of skin appendage morphogenesis. The absence of ACC in K14Cre+Kctd1fl/flKctd15fl/fl mice also suggests that ACC is a consequence of the loss of these complexes in cells other than keratinocytes.

LOSS OF KCTD1/KCTD15 FUNCTION IN CRANIAL NCCS LEADS TO ACC

Clinical analysis shows that the majority of membranous ACC occurs over midline cranial sutures and can be associated with underlying bone/suture defects and heterotopic meningeal/brain tissue (Bessis et al., 2017). Mouse studies determined that the midline cranial sutures (interfrontal and sagittal), and several midline skull bones (including nasal, frontal, and interparietal bones), are derived from cranial NCCs (Mishina and Snider, 2014). The observation that ACC forms at sites where the underlying bone/suture structures are derived from NCC populations suggests that ACC is a secondary consequence of a defect or absence in NCC-derived midline cranial suture cells and is not due to a primary keratinocyte abnormality. Indeed, inactivation of KCTD1/KCTD15 complexes in NCCs (in Wnt1Cre+Kctd1fl/flKctd15fl/fl mice), but not of Kctd1 or Kctd15 alone, resulted in congenital bone/suture defects of NCC-derived structures of the midline skull associated with overlying membranous ACC-like skin defects with epidermal thinning (Raymundo et al., 2023). Newborn Wnt1Cre+Kctd1fl/flKctd15fl/fl mice had a widening of midline cranial sutures with a receded osteogenic front, coupled with shortened frontal bones. The ACC-like lesions in these mice occurred atop these abnormal interfrontal or sagittal sutures and had a flattened epidermis with a thin basal layer, resembling membranous ACC (Raymundo et al., 2023). Analysis of scRNA-seq data of interfrontal suture cell populations at E16.5 or E18.5 demonstrated that these NCC-derived mesenchymal cells and adjacent osteogenic front cells normally express multiple growth factors known to induce keratinocyte proliferation and migration (Holmes et al., 2020; Raymundo et al., 2023), such as keratinocyte growth factor (Kgf/Fgf7), Igf1, Igf2, and Fgf10 (Gunschmann et al., 2013; Seeger and Paller, 2015). Receptor/ligand interaction analyses of these scRNA-seq data suggest that NCC-derived cranial sutures act as signaling hubs that orchestrate the formation of the overlying skin: incoming signals into the hypodermis from NCC-derived interfrontal cranial suture mesenchymal cells included IGF, EGF, and FGF families of growth factors as well as other growth factors or signaling molecules (Raymundo et al., 2023; Figure 2). Whether a single specific growth factor or the synergistic effects of multiple such factors secreted by NCC-derived cranial suture mesenchymal cells and osteogenic front cells are critical for the formation of the overlying scalp skin remains to be determined. Igf1 is a likely candidate for an inductive signaling factor that has an important role in midline scalp skin formation, as Igf1 is particularly highly expressed by interfrontal suture cells and because epidermal Igf1r deficiency results in epidermal thinning (Gunschmann et al., 2013) as observed in ACC-like lesions in Wnt1Cre+Kctd1fl/flKctd15fl/fl mice. It remains to be determined whether the loss of KCTD1/KCTD15 function in cranial NCCs results in these midline suture defects and ACC due to abnormal differentiation of those NCC-derived mesenchymal cells, whether these cells are diminished in number due to a cell migration defect, or whether their function in the suture tissue microenvironment is impaired.

Figure 2: Proposed model for the role of NCC-derived cranial suture cells in scalp skin morphogenesis.

The observation of ACC in mice that lack KCTD1/KCTD15 complexes in NCCs but not in those that lack KCTD1/KCTD15 in keratinocytes suggests that NCC-derived cells of the midline cranial sutures induce the formation of the overlying scalp skin by secreting inductive growth-promoting signals (e.g., keratinocyte growth factors) to the overlying keratinocytes that migrate and proliferate to form the overlying scalp skin (yellow arrows).

Notably, inactivation of KCTD15 or KCTD1/KCTD15 in NCCs resulted in a white belly patch, a pigmentation defect that suggests abnormal migration of NCCs/NCC-derived melanoblasts towards the abdominal midline. This is similar to defects observed in mouse mutants with impaired function of transcription factors Pax3 or AP-2α, which are key regulators of NCC function (Brewer et al., 2002; Johnson et al., 2020; Olaopa et al., 2011). KCTD15 is more highly expressed than KCTD1 in melanocytes of the developing mouse skin, and the size of the abdominal pigmentation defect increases in size concomitantly with the reduction of overall Kctd1/Kctd15 transcript levels, consistent with an overlapping function and partial compensation effects between KCTD1 and KCTD15 in these cells (Raymundo et al., 2023). This finding may also suggest that loss of KCTD1/KCTD15 impairs the migration of NCCs and results locally in a diminished number of NCC-derived mesenchymal cells in midline cranial sutures. However, it is currently unknown through which downstream signaling mechanisms loss of KCTD1/KCTD15 impairs cell function or differentiation of NCC-derived cells to result in the observed phenotypes in Wnt1Cre+Kctd1fl/flKctd15fl/fl mice or in K14Cre+Kctd1fl/flKctd15fl/fl mice.

LOSS OF KCTD1/KCTD15 FUNCTION IN NCCS EXPLAINS ADDITIONAL CRANIOFACIAL PHENOTYPES SEEN IN SEN SYNDROME

SEN syndrome patients or individuals with KCTD15 missense mutations can also have an abnormal nose (small nasal tip, anteverted nares) and anosmia. Nasal bones are derived from NCCs and newborn Wnt1Cre+Kctd1fl/flKctd15fl/fl mice displayed an almost complete loss of nasal bones and also diminished nasal cartilage (Raymundo et al., 2023). Thus, the nasal abnormalities in these patients are a consequence of impaired NCC-derived cells that form the nasal bone/cartilage structures.

SEN syndrome patients also often have tooth abnormalities, in some cases with an absence of incisors (Marneros et al., 2013). Interactions between the dental epithelium and the underlying NCC-derived mesenchymal cells of the dental papilla are pivotal for tooth morphogenesis. Lack of Kctd1 and Kctd15 in the dental epithelium (in K14Cre+Kctd1fl/flKctd15fl/fl mice) did not affect incisor formation, whereas their absence in the NCC lineage (and thus in the dental papilla) in Wnt1Cre+Kctd1fl/flKctd15fl/fl mice resulted in an absence of mandibular and maxillary incisors (Raymundo et al., 2023). The abnormal nose and absence of incisors in Wnt1Cre+Kctd1fl/flKctd15fl/fl mice further underscore that KCTD1/KCTD15 complexes are critical regulators of NCC-derived cells that differentiate into craniofacial structures.

DISTINCT TISSUE-SPECIFIC EXPRESSION PATTERN OF KCTD1 AND KCTD15 EXPLAINS NON-OVERLAPPING PHENOTYPES DUE TO KCTD1 VERSUS KCTD15 MUTATIONS

The high expression of KCTD1 but not KCTD15 in DCTs of the kidney correlated with the observation of kidney abnormalities and progressive renal failure in SEN syndrome patients with dominant-negative KCTD1 mutations, whereas patients with KCTD15 mutations did not have these kidney abnormalities (Figure 3). Notably, the two individuals with the KCTD15D104H mutation were also found to have congenital cardiac abnormalities: tetralogy of Fallot and patent ductus arteriosus (Miller, 2024; Miller, 2018). Similarly, Kctd15 null mice developed congenital cardiac defects: perimembranous ventricular septal defects and bicuspid aortic valves (Raymundo et al., 2023). Cardiac NCCs contribute to the development of these cardiac structures, and the same abnormalities were also found in Wnt1Cre+Kctd1fl/flKctd15fl/fl mice (but not in Wnt1Cre+Kctd15fl/fl mice or Kctd1 null mice). ScRNA-seq data analysis of embryonic mouse hearts shows that Kctd15 is earlier expressed than Kctd1 in cells of the cardiac outflow tract and the atrioventricular canal (~E7.75), whereas both are expressed in cardiac NCCs at a later stage (~E8.25-E9.25; Raymundo et al., 2023). This suggests that Kctd15 expression in cells forming the cardiac outflow tract and the atrioventricular canal is critical for the proper development of the perimembranous ventricular septum and the aortic valve, a function that cannot be compensated for by Kctd1. In cardiac NCCs, KCTD1/KCTD15 complexes contribute to the formation of these cardiac structures and KCTD1 and KCTD15 can compensate for each other’s loss. These data uncover a previously unknown critical role of KCTD15 in the development of the perimembranous ventricular septum and the aortic valve (Figure 3).

Figure 3: Spectrum of clinical manifestations due to lack of KCTD1 and/or KCTD15 function can be explained by the cell type-specific expression pattern of KCTD1 versus KCTD15.

In DCTs of the kidney, KCTD1 but not KCTD15 is strongly expressed, explaining DCT defects in Kctd1 null mice and in SEN syndrome patients with dominant-negative KCTD1 mutations. KCTD15 loss cannot be compensated for by KCTD1 in cells that give rise to the perimembranous ventricular septum and aortic valve. In keratinocytes, KCTD1 and KCTD15 can compensate for each other’s loss, and only combined loss of KCTD1 and KCTD15 function results in skin appendage defects (impaired morphogenesis of hair follicles, sebaceous glands, and eccrine sweat glands). Loss of KCTD1/KCTD15 function in cranial NCCs leads to midline skull defects and ACC, loss of incisors, and nose abnormalities. CTD: C-terminal domain; BTB: BTB domain.

SUMMARY

The identification of dominant-negative mutations in KCTD1 or KCTD15 in individuals who develop ACC provides direct genetic evidence for a key role of KCTD1/KCTD15 complexes for the proper formation of the midline scalp skin during development. The generation of mice that lack Kctd1 and Kctd15 in keratinocytes versus in NCCs demonstrates that the epidermal defect in ACC is not a consequence of lack of KCTD1/KCTD15 function in keratinocytes but instead a secondary consequence of abnormalities of NCC-derived midline cranial suture cells due to loss of KCTD1/KCTD15 complexes in these cells. In contrast, KCTD1/KCTD15 complexes in keratinocytes are required for the proper development of hair follicles, sebaceous glands, and eccrine sweat glands. Notably, cell type–specific inactivation of Kctd1 and/or Kctd15 in genetic mouse models also links gene expression differences of KCTD1 and KCTD15 in various cell types to the observed distinct clinical phenotypes in patients with KCTD1 versus KCTD15 mutations (Figure 3). For example, these data reveal that KCTD1 loss in DCTs of the kidney cannot be compensated for by KCTD15, whereas KCTD15 function in cells giving rise to structures derived from the cardiac outflow tract/atrioventricular canal cannot be compensated for by KCTD1. The remarkable similarities of the clinical features between individuals with KCTD1 or KCTD15 mutations with the observed phenotypes in the various cell type-specific Kctd1/Kctd15 mouse mutants establish the clinical relevance of the findings in these mice. Future studies need to delineate through which downstream molecular mechanisms the loss of KCTD1/KCTD15 complexes impair the cellular functions that lead to the observed phenotypes, including ACC. A potential pathomechanism to consider includes an overall hyperactivation of AP-2 transcription factor activity as a consequence of loss of KCTD1/KCTD15, since both KCTD1 and KCTD15 act as inhibitors of AP-2 and because a downregulation of AP-2 activity has been suggested to be required for keratinocyte maturation (Li et al., 2019). Other signaling pathways that may be altered as a consequence of KCTD1/KCTD15 deficiency and that may contribute to the observed phenotypes include changes in Shh or Wnt/β-catenin signaling activities. The identification of KCTD1 and KCTD15 mutations in patients and the elucidation of the roles of KCTD1/KCTD15 complexes in genetic mouse models serve as an example of how uncovering the genetic basis of rare genodermatoses can lead to fundamental new insights into mechanisms that control skin development and skin appendage morphogenesis. These discoveries open up an exciting new area of investigation with broad significance for skin biology.