Adipose and Hair Function: An aPPARent Connection

Abstract

Adipose tissue plays essential roles in various aspects of skin physiology, from regulating hair follicle morphogenesis to wound healing. Peroxisome proliferator-activated receptor gamma is important for the maintenance of adipose tissue and has been implicated in some types of hair loss; however, its function during the hair cycle is still unclear. Sardella et al. investigate the role of peroxisome proliferator-activated receptor gamma in hair follicle morphogenesis using a novel global peroxisome proliferator-activated receptor gamma-null mouse.

Adipose tissue, which consists of a heterogeneous population of cells, is critical for metabolic and tissue homeostasis. In mammalian skin, dermal adipocytes regulate hair growth, dictate tissue remodeling during wound healing, and are vital for skin barrier and antimicrobial host defense function. During the anagen phase of hair cycling, increased skin thickness is partially mediated by hyperplasia of intradermal adipocytes, known as adipogenesis. This process has recently been shown to be dependent on platelet-derived growth factor signaling (Festa et al., 2011; Rivera-Gonzalez et al., 2016). Adipogenesis requires the activation and nuclear translocation of the transcriptional regulator peroxisome proliferator-activated receptor gamma (PPARγ), which allows preadipocytes to differentiate into mature, lipid-filled adipocytes (Figure 1). In human patho-biology, defective PPARγ signaling is proposed to contribute to primary cicatricial or scarring alopecia, particularly, lichen planopilaris (Karnik et al., 2009). Sardella et al. (2017) further elucidate the regulatory role of adipose tissue during hair follicle (HF) morpho-genesis in the context of PPARγ using a clever murine model with an epiblastspecific deletion of the PPARγ allele (Pparγ^Δ/Δ), allowing for otherwise lethal global deletion of PPARγ.

Figure 1.

The model as proposed by Sardella et al. (2017). PPARγ is crucial for regulating adipogenesis, during which preadipocytes differentiate into mature adipocytes. Additionally, PPARγ promotes sebocyte maturation and proper hair follicle morphogenesis and maintenance of the pilosebaceous unit. Dysregulation or ablation of PPARγ signaling results in a temporal delay of hair follicle morphogenesis and and defective hair cycling.

PPARs (PPARα, γ, and β/δ) are members of the nuclear receptor superfamily of transcription factors that regulate the expression of genes involved in inflammation and lipid homeostasis and are activated by lipophilic ligands. As essential regulators of gene expression and differentiation, PPARs display unique expression in various vertebrate tissues including skin, sebaceous glands (SGs), and lymphoid tissue. Previous studies have shown that while PPARδ is the most abundant PPAR in the adult epidermis, PPARγ is the most highly expressed PPAR in sebocytes. PPARγ exclusively initiates the differentiation of sebocytes in the SG, while PPARδ is thought to induce sebocyte maturation (Rosenfield et al., 1999). Previous studies have shown that activation of PPARγ stimulates keratinocyte differentiation, improves permeability barrier homeostasis, and stimulates epidermal lipid synthesis (Mao-Qiang et al., 2004). The activation of all three PPARs trigger anti-inflammatory responses by modulating various immune effectors and adhesion molecules. As such, the vast regulatory functions of PPARγ in lipid metabolism and inflammation imply its vital role in the maintenance of the pilosebaceous unit (PSU).

Several approaches have been adopted to study the in vivo functions of PPARγ using tissue-specific deletions. To study the role of PPARγ in lichen planopilaris, Karnik et al. (2009) generated a mouse model with a targeted deletion of PPARγ in bulge HF stem cells (HFSC) using keratin 15 promoter-driven recombination. The study of PPARγ function by whole body knockout in vivo is not possible because homozygous deletion of PPARγ is embryonic lethal due to defective placental function. Sardella et al. (2017) circumvented this issue by utilizing a whole-body PPARγ knockout mouse that retains PPARγ expression in the trophoblast, previously described by Nadra et al. (2010). In theory, the whole-body deletion of PPARγ using epiblast-specific knockout mice provides a great opportunity to explore the physiological consequences of global PPARγ loss-of-function in adult mice. The major limitation of this approach is that it is difficult to determine primary and cell-intrinsic effects due to multi-system involvement. The alopecia and perifollicular inflammation in the K15-Cre;PPARγ^fl/fl mice (Karnik et al., 2009) were recapitulated in the Pparγ^Δ/Δ mice (Sardella et al., 2017); however, the latter exhibited a complete absence of SGs and total lipoatrophy, whereas the former exhibited SG atrophy. Most importantly, Karnik et al. (2009) showed normal hair and skin postnatally until 3 months when the mice exhibited progressive hair loss, whereas Sardella et al. (2017) showed transiently delayed HF morphogenesis postnatally, slower HF cycle, shorter hair, and progressive hair loss around 6 months. The similar inflammation and progressive phenotype can be explained by SG dystrophy in both models, and the difference in HF morphogenesis can be attributed to adipose tissue.

Sardella et al. (2017) observed that Pparγ^Δ/Δ mice not only exhibit delayed HF morphogenesis, but also sustained delayed entry into the first and subsequent HF cycles (Figure 1). Various HFSC and hair differentiation markers were significantly downregulated in Pparγ^Δ/Δ mice, suggesting defective HF stem cell function. To address the question of whether or not lipodys-trophy is directly responsible for these effects, the authors used a PPARγ-independent lipodystrophy model (AZIP^tg/+) and generated a fat-specific PPARγ null mouse (PparγF^Δ/Δ). Both Pparγ^Δ/Δ and AZIP^tg/+ mice lack mature adipocytes while retaining normal SGs and adipocyte progenitor cells. Interestingly, both PparγF^Δ/Δ and AZIP^tg/+ mice exhibited delayed HF morphogenesis, as well as reduced expression of several hair morphogenetic markers, similar to Pparγ^Δ/Δ mice. Although Pparγ^Δ/Δ and AZIP^tg/+ mice exhibited defective entry into the first HF cycle (catagen), adult (6-month-old) AZIP^tg/+ mice exhibited normal HF cycling compared with Pparγ^Δ/Δ mice. In addition, differential gene expression analysis showed significant down-regulation of adipocyte maturation and HF morphogenesis markers in all murine lipodystrophy models. To show that adipose tissue is required for HF morphogenesis, Sardella et al. (2017) engrafted skin from Pparγ^Δ/Δ, AZIP^tg/+, and control mice onto athymic nude mice. Remarkably, both lipodystrophic engraftments completely rescued defective HF morphogenesis and upregulated hair morphogenetic markers. In an important contrast, Pparγ^Δ/Δ grafts ultimately displayed increased inflammation and macrophage infiltration unlike AZIP^tg/+ grafts. Taken together, these findings provide strong evidence that adipocytes are required for the normal regulation of early HF morphogenesis and that dysfunctional PPARγ signaling can have negative implications on late HF morphology and cycling due to disrupted HFSC regulation, abolishment of SGs, and increased skin inflammation.

The results from this study highlight the importance of adipose tissue in temporally regulating the HF cycle during morphogenesis. Although the authors link PPARγ dysfunction to the aberrant homeostasis of the PSU, future work will further explore whether or not the lack of SGs alone is responsible for the defective hair cycling observed in Pparγ^Δ/Δ mice and the underlying molecular and cellular mechanisms. The development of SGs and the dynamics of stem and progenitor cell compartmentalization during PSU morphogenesis are incompletely defined. Lrig1 has been identified as a key HFSC marker that gives rise to mature sebocytes and is responsible for the maintenance of the PSU. In addition, canonical Wnt/bcatenin signaling has been strongly implicated in epidermal cell fate decisions, including regulating sebaceous lineage differentiation—specifically, antagonism of canonical Wnt signaling promotes sebocyte differentiation. Whether or not PPARγ is responsible for defining the HFSC reprogramming landscape by regulating cell fate toward sebocyte lineages is an interesting unresolved question.

This report also raises other interesting questions. Why does chronic inflammation occur during late hair cycling, but not during early postnatal HF development in Pparγ^Δ/Δ mice? Karnik et al. (2009) showed that the loss of PPARγ signaling results in the accumulation of proinflammatory lipids, such as prostaglandins, generated by PTGS2 (COX-2). Prostaglandins are known to regulate the hair cycle. Garza et al. (2012) demonstrated that prostaglandin D2 inhibits hair growth and is elevated in the bald scalps of males with androgenetic alopecia. In contrast, PGF_2a and PGE₂ have been shown to stimulate hair growth. Wan et al. (2007) showed that targeted deletion of maternal PPARγ results in the production of milk with elevated level inflammatory lipids that cause suckling neonates to develop alopecia and tissue inflammation. Nursing neonates had elevated levels of COX-1, COX-2, and prostaglandin transport in their skin. Aspirin, a COX-1/2 inhibitor, prevented alopecia in those neonates. Similar metabolite increases in the eicosanoid pathway were also observed by Karnik et al. (2009). Are prostaglandins the intermediary modulators of the observed inflammatory infiltrate (macrophages, neutrophils) and delayed HF morphogenesis due to the absence of PPARγ? Although this study provides strong in vivo evidence, future studies should test, in vitro, the notion that PPARγ is directly responsible for the downregulation of key morphogenetic markers that lead to delayed HF morphogenesis and inflammation. Perhaps in the future, the immune pathway should be dissected in vitro with individual and co-culture experiments to show how direct or indirect PPARγ ablation leads to inflammation.

The clinical implications of this article are important to consider given that abnormal PPARγ has been associated with the cicatricial alopecia lichen planopilaris. One current model is that progressive abrogation of PPARs leads to proinflammatory infiltrate and lipid accumulation, and permanent destruction of the PSU. It has been proposed that defective SG function is a pathological cause for cicatricial alopecia. This is based, in part, on observations of the asebia mutant mouse (Scd1ab), which displays a scarring alopecia phenotype (Sundberg et al., 2000) due to a lack of the lipid-metabolizing enzyme steroyl-CoA desaturate 1 (SCD1). This results in SG atrophy, defective sebum production, and chronic skin inflammation due to mast cell infiltration. The asebia mouse shares similarities with the Pparγ^Δ/Δ mouse in this study in that both mice exhibit SG atrophy, dysregulation of lipid metabolism, proinflammatory events, and alopecia. It is worth noting that Sardella et al. (2017) have not shown histology consistent with end-stage scarring alopecia in Pparγ^Δ/Δ mice: total deletion of the entire PSU in wide swaths of skin, although this may exist. Therefore, more work is required to determine if the Pparγ^Δ/Δ mouse is indeed a valid model for scarring alopecia.

Altogether, Sardella et al. (2017) provide insights into PPARγ and its role in timing HF morphogenesis and cycling. Future work will delineate the exact mechanism by which PPARγ regulates perifollicular inflammation, and its potential relevance to human alopecias.

Clinical Implications.

• Lipoatrophy causes a temporal delay of postnatal hair follicle morphogenesis and differentiation.

• Loss of peroxisome proliferator-activated receptor gamma causes sebaceous gland atrophy, lipoatrophy, and perifollicular inflammation resulting in delayed hair follicle morphogenesis and alopecia.

• Future therapeutic strategies for scarring alopecias will not only need to focus on peroxisome proliferator-activated receptor gamma stimulation, but also other underlying causes for inflammation.