The Black-Box Illuminated: Signals and Signaling

Abstract

Unraveling the signaling pathways that transmit information from the cell surface to the nucleus has been a major accomplishment of modern cell and molecular biology. The benefit to man is seen in the multitude of new therapeutics based on the illumination of these pathways. While considerable insight has been gained in understanding homeostatic and pathological signaling in the epidermis and other skin compartments, the translation into therapy has been lacking. This review will outline advances made in understanding fundamental signaling in several of the most prominent pathways that control cutaneous development, cell fate decisions, and keratinocyte growth and differentiation with the anticipation that this insight will contribute to new treatments for troubling skin diseases.

Calcium Signaling in Keratinocytes

It has been 3 decades since awareness of the unique role of calcium in the regulation of growth and differentiation of keratinocytes first came to light through studies of cultured keratinocytes (Hennings et al., 1980). Since that time modulation of calcium in vivo and in vitro has been the major tool used to illuminate the fine structure of keratinocyte and epidermal biology and has contributed to understanding the molecular basis of several skin diseases. Beyond keratinocytes, calcium is increasingly recognized as a central transmitter of signals in all cells, and calcium signaling is dynamically controlled during normal cell cycles and in resting states (Dupont et al., 2011; Putney, 2009; Roderick and Cook, 2008). The central importance of calcium in cell physiology is clearly demonstrated by its complex regulation involving channels, pumps, sensors, binding proteins, hormones, and receptors both on the plasma membrane and intracellular organelles. Furthermore in both excitable and non-excitable cells there is a constant flux of calcium exchanged from intracellular compartments and across the plasma membrane, a process termed calcium oscillations. Under differing conditions the cytosolic free calcium can range from 100nM to 1μM and return to equilibrium may occur in seconds, minutes or hours depending on the nature of the stimulus and the requirements of the functional response. The plasma membrane of most cells is inhabited by a variety of channels for the influx of calcium from the extracellular space (Figure 1). Among these are store operated channels (SOCE) that activate influx in response to depletion of intracellular stores. Proteins known to be associated with this pathway include STIMs that monitor calcium content of endoplasmic reticulum (ER) stores. Depletion of intracellular stores is sensed by STIMs that then translocate to the plasma membrane and interact with Orai, the pore forming unit of the channel and TRPC (transient receptor potential C) to stimulate calcium influx. Additional influx is regulated by second messenger operated channels (SMOC) responsive to diacylglycerol, receptor operated channels responsive to hormones (ROC) and voltage gated channels (VGCC). Calcium influx is also downstream from receptor tyrosine kinases including EGFR. ATP dependent calcium pumps reside on the plasma membrane and the membranes of intracellular storage sites such as ER, golgi and mitochondria. These serve to pump out excess cytosolic calcium through the plasma membrane (PMCA, NCX) or into storage sites (SERCA) where calcium remains bound to high capacity calcium storage proteins such as calreticulin of the ER. Of particular importance in calcium signaling are G-protein coupled receptors, including the calcium-sensing receptor (CaR) on the plasma membrane, that activates membrane bound phospholipaseC to generate inositol phosphates, particularly inositol 1,4, 5 trisphosphate (IP3) that stimulate receptors on intracellular organelles to release calcium stores. This elevation of intracellular free calcium is translated into functional responses through calmodulin and other downstream effectors. What has become apparent in the last 3 decades is that all of these components of calcium signaling are major regulators of keratinocyte biology.

Figure 1. Integration of the calcium signaling circuitry:

The major regulators of calcium homeostasis in keratinocytes are depicted. Plasma membrane pumps and channels (PMCA, NCX, SOCE) regulate flux in and out of the cytosol. G protein coupled receptors (CaR and others not shown) initiate signals that modify compartmentalized calcium stores (e.g. IP3). Calcium ATPases on organelles (SERCA, SPCA1) monitor and replenish intracellular storage sites. The figure is modified from (Savignac et al, 2011) with permission from the publisher.

The calcium gradient inside and out

For almost 25 years skin biologists have known that the avascular intact epidermis maintains a calcium gradient that is lower in the basal compartment and enriched in granular cells before a steep drop off in the stratum corneum (Elias et al., 2002). Disturbance of this gradient by barrier dysfunction or other means prevents normal keratinocyte differentiation and accelerates lamellar body secretion. While several methods available to early investigators using fixed tissues confirmed the existence of the gradient, newer techniques in living tissues suggest that the variation in the strata arise from differences in intracellular calcium stores and variations exist within populations in the basal cell compartment (Behne et al., 2011; Celli et al., 2010). This is not surprising as it is well documented that graded levels of extracellular calcium elicit a graded differentiation response in keratinocytes (Yuspa et al., 1989), buffering of intracellular calcium prevents terminal differentiation of keratinocytes (Li et al., 1995a) and the expression of early and late markers of differentiation are regulated by different intracellular calcium compartments (Li et al., 1995b). How could these compartmental changes be physiologically regulated? Elevation of extracellular calcium activates several second messenger systems. In particular, elevated calcium activates phospholipase Cγ1 and δ1 to increase inositol lipids and catalyze the release of calcium from intracellular stores, acutely increasing cytosolic calcium (Jaken and Yuspa, 1988; Lee and Yuspa, 1991; Punnonen et al., 1993). What follows is not completely understood but involves CaR, SOCE, TRPC and other plasma membrane cation channels (Denda et al., 2006; Mauro et al., 1997) to sustain the rise in intracellular calcium and restore calcium levels in storage sites. Interference with CaR expression blocks both differentiation and adherens junction formation in human keratinocytes, and keratinocyte differentiation is defective in CaR null mice (Tu et al., 2008; Tu et al., 2004).

Downstream effectors of calcium signaling

The observation that keratinocytes from all species tested proliferate actively in calcium below 0.1mM while most other cells growth arrest or die suggests an evolutionary function is conserved. Barrier formation and specialized cell adhesions are obvious unique functions for keratinocytes, and calcium is essential for crosslinking cornified envelopes, desmosome assembly and creation of adherens junctions. Furthermore depletion of ER calcium by barrier disruption or chemical stress results in rapid secretion of lamellar bodies in the stratum granulosum to repair the damage (Celli et al., 2010). The importance of assembling these structures sequentially and repairing the barrier is clear and could account for compartmental regulation of calcium signaling. Consistent with an evolutionary view is the remarkable clustering of genes essential for terminal differentiation together with calcium binding S100 proteins on human chromosome 1q21. Beyond the barrier what are the downstream effectors of calcium signaling that carry out the other messages? Activation of calcium dependent protein kinase C α is one effector linked closely to transcriptional regulation of spinous and granular cell proteins through AP-1 activity (Denning et al., 1995; Rutberg et al., 1997). Furthermore, genes expressed during keratinocyte differentiation contain calcium dependent regulatory elements (Rothnagel et al., 1993). The calcium-calmodulin response pathway modifies a number of pathways regulating cell behavior (Wayman et al., 2011). Among the proteins activated through calcium-calmodulin is the serine phosphatase calcineurin. In keratinocytes and other cells calcineurin dephosphorylates NFAT (nuclear factors of activated T cells), allowing NFAT to enter the nucleus and regulate keratinocyte proliferation and stem cell quiescence through p21 in conjunction with Notch signaling (Dotto, 2011). Recent studies have implicated TRPV1, TRPV3 and TRPA1 calcium channels that are highly expressed in keratinocytes to also contribute to epidermal differentiation and inflammation in addition to sensory responses (Cheng et al., 2010; Toth et al., 2009). A compelling role for calcium signaling in keratinocyte homeostasis comes with the discovery of inactivating mutations in SERCA2 (ATP2A2), the endoplasmic reticulum calcium ATPase in Darier’s disease, and mutations in ATP2C1 (SPCA1), the Golgi calcium ATPase in Hailey-Hailey disease (Savignac et al., 2011; Sudbrak et al., 2000). In both cases the mutation results in calcium depletion in the organelles and disturbances of skin barrier and keratinocyte adhesion and differentiation. It is likely that future studies will reveal additional cutaneous pathology as a consequence of altered calcium homeostasis.

C-ing the Forest for the Trees: Phospholipase C and Protein Kinase C Signaling

Early Explorations

The discovery in the early 1980s that protein kinase C (PKC) was the primary receptor for tumor promoting phorbol esters used in mouse skin chemical carcinogenesis studies captured the imagination of skin carcinogenesis researchers (Nishizuka, 1984). This revelation identified the phorbol ester receptor as a central molecule in phospholipase C (PLC)-coupled growth factor receptor signaling and promised to dramatically simplify and focus mechanistic studies on the promotion stage of mouse skin DMBA/TPA chemical carcinogenesis. From this early perspective, phorbol esters were simply substituting for chronic mitogenic growth factor stimulation and thus directly driving proliferation of keratinocytes. Almost three decades of intensive research on PLC-PKC signaling in cutaneous biology have unearthed multiple criss-crossing pathways involving lipid-derived messengers, phosphorylation and protein interactions, and have expanded from skin carcinogenesis to epidermal differentiation, wound healing and inflammation (Figure 2).

Figure 2. Integration of PKC signaling circuitry:

Protein kinase C isoforms can be activated by the second messenger inputs calcium (Ca²⁺) and DAG generated by the activity of PLC. Intracellular calcium activates PKCα, β and γ via their C2 domains, while DAG and TPA activate PKCα, β, γ, δ, ε, η and θ via C1 domains. Src family kinases can also tyrosine phosphorylate and alter the expression of PKCδ. The outputs of PKC signaling are isoform dependent as indicated: PKCα (differentiation-associated growth arrest and cytokine-mediated inflammation), PKCβ (melanogenesis), PKCδ (squamous differentiation, growth arrest and apoptosis), PKCη (squamous differentiation and growth arrest), PKCε (hyperplasia).

The basic architecture of PLC/PKC signaling consists of PLC activation via coupling to either G protein-coupled receptors or receptor tyrosine kinases, resulting in the hydrolysis of phosphatidylinositol 4,5-bisphosphate (PIP₂) (Suh et al., 2008). The two second messengers produced by this cleavage are diacylglycerol (DAG), which binds C1 domains and activates PKC, and inositol 1,4,5-trisphosphate (IP₃), which binds to the IP₃ receptor and triggers the release of Ca²⁺ from intracellular stores. Phorbol esters, such as 12-O-tetradecanoyl-phorbol-13-acetate (TPA), are potent long-lived diacylglycerol mimetics which have strong agonist activity, but at high dose trigger proteasome-dependent and independent degradation of PKC resulting in a type of receptor desensitization (Leontieva and Black, 2004).

Signaling Diversification

The discovery and characterization of multiple genes encoding both PLC and PKC have been major driving forces in the mapping of their interconnected signaling mechanisms. The PLC gene family has 13 members (β1, β2, β3, β4, γ1, γ2, δ1, δ3, δ4, ε, η1, η2, ζ) and is more diverse than the 9 member PKC family (α, β, γ, δ, ε, η, θ, ζ, ι) (Parker and Murray-Rust, 2004; Suh et al., 2008). Each gene family has subfamilies defined by functional domain composition. These domains dictate the activation mechanisms and specificity, as well as the effector functions. For example, the classical, Ca²⁺-responsive PKC isoforms (α, β, γ) are responsive to both DAG/TPA by virtue of their C1 domains, and to Ca²⁺ by virtue of their C2 domains. Other PKC subfamilies lack canonical C2 domains (δ, ε, η, θ) and C1 domains (ζ, ι) influencing the cofactor requirements and kinetics of activation (Lenz et al., 2002). Effector specificity is determined primarily by selective substrate access dictated by scaffolding proteins which bind to unique targeting sequences in each PKC isoform (Kheifets and Mochly-Rosen, 2007). Thus, despite PKC isoforms having remarkably similar kinase substrate specificity, they mediate very different functions in the cell due to unique subcellular localization, cellular substrate phosphorylation, and activation mechanisms.

The PLC/PKC signaling modules relevant to normal skin biology involve primarily terminal differentiation. PKCδ and PKCη are linked to keratinocyte squamous differentiation (Adhikary et al., 2010; Denning, 2004). For calcium-induced keratinocyte differentiation, the G protein-coupled calcium sensing receptor (CaR) and PLC-γ1 are involved, and mechanisms beyond raising intracellular calcium may contribute to the pro-differentiation response and PKC signaling (Tu et al., 2004; Xie and Bikle, 2007). PKC is also linked to differentiation functions of the pigment-producing melanocyte (Park et al., 2004; Park et al., 1999). PKCβ is able to phosphorylate and activate tyrosinase, the rate limiting enzyme in melanin biosynthesis and is thus implicated in melanogenesis.

The heterogeneic PKC signaling landscape has required cancer biologists to integrate PKC signaling into the larger cellular and tissue framework to understand how phorbol esters promote tumor development. In mouse skin chemical carcinogenesis, initiating H-Ras mutations result in increase phosphatidylinositol hydrolysis, thus elevating DAG levels and activating PKCα (Denning et al., 1995; Lee and Yuspa, 1991). H-Ras activation also elevates Src family kinase signaling, resulting in inactivation of the pro-apoptotic PKCδ through tyrosine phosphorylation and/or transcriptional silencing (D’Costa et al., 2006; Denning et al., 1993; Geiges et al., 1995). Chronic TPA treatment results in selective survival and proliferation of H-Ras mutant keratinocytes, but this occurs in the context of general epidermal hyperplasia and cytokine release by keratinocytes in response to TPA. PKCα, PKCε and NF-κB have been implicated in this cytokine and inflammatory response, but precisely how the combined effects of H-Ras activation and chronic PKC activation results in squamous skin cancers is still unclear (Cataisson et al., 2005; Wheeler et al., 2005). A similar scenario exists for UV skin carcinogenesis as UV elevates DAG levels, pro-apoptotic PKCδ expression is repressed, and chronic activation of PKCε promotes regenerative hyperplasia (Aziz et al., 2007; Jansen et al., 2001; Punnonen and Yuspa, 1992). One reasonable model for the role of PKC isoforms in skin carcinogenesis is selective activation of PKCα driving inflammation while PKCε drives cell proliferation. In association, inactivation/repression of pro-apoptotic PKCδ promotes cell survival. This model is supported by studies in transgenic mice in which PKCε transgenic mice are highly susceptible to both chemical and UV skin carcinogenesis while PKCδ mice are resistant to skin carcinogenesis (Jansen et al., 2001; Reddig et al., 1999; Reddig et al., 2000).

Non-Standard Keratinocyte Responses

Several of the fundamental signaling molecules related to PLC/PKC signaling function very differently in epidermal keratinocytes compared with most other cell types. Central among these is calcium, which is abbreviated as “C” in protein kinase C despite most PKC isoforms lacking a canonical C2 domain and being calcium-independent. As dicussed previously, most cells bathed in serum in vivo or grown in tissue culture proliferate normally in >1 mM calcium. In contrast, epidermal keratinocytes proliferate well in <0.1 mM calcium and are induced to undergo squamous differentiation by >1 mM Ca²⁺ in association with PLC/PKC activation(Denning et al., 1995; Hennings et al., 1980). Keratinocytes also express PKCη, a PKC isoform unique to squamous differentiation tissues (Kashiwagi et al., 2002). The function of the PKC effector NF-κB in keratinocytes is also unusual. NF-κB is pro-survival and thus promotes cancer development in most cell types, however the NF-κB upstream activating kinase IKKα promotes keratinocyte differentiation, and blockade of NF-κB activation with a dominant/negative IκB is a potent oncogene in human keratinocytes (Dajee et al., 2003; Hu et al., 2001). The signaling responsible for these unusual keratinocyte responses is still enigmatic, but is an area of active investigation.

Therapeutics

Despite the complexity of PLC/PKC signaling and lack of isoform selective small molecule activators/inhibitors, the diverse and dramatic effects of PLC/PKC modulation on skin biology have attracted much interest from drug developers. Ingenol-3-angelate (ingenol mebutate, PEP005), which activates PKCδ and has unusually deep tissue penetration, has promising clinical activity against premalignant actinic keratosis (Hampson et al., 2005; Li et al., 2010; Siller et al., 2009). Another area of potential therapeutics is inhibition of PKCβ to block pigmentation. Increased knowledge of PKC activation mechanisms, effector specifics and the non-standard keratinocyte signaling responses will undoubtedly continue to challenge the imagination of investigative dermatologists for at least another 25 years.

The Profound Influence of EGFR Signaling in Cutaneous Biology

Overview of the EGFR signaling pathway and relevance in skin biology

The Epidermal Growth Factor Receptor (EGFR) is expressed almost ubiquitously with the exception of mature hematopoietic cells. EGFR is a member of a family of four receptors namely ERBB1 or EGFR, ERBB2, ERBB3 and ERBB4. Ligand binding triggers receptor dimerization followed by transphosphorylation on multiple tyrosine residues (Schulze et al., 2005). Homo or heterodimer receptor partners interact with a growing family of ligands like Epidermal Growth Factor (EGF), Transforming Growth Factor alpha (TGF-alpha), Heparin-Binding EGF-like Growth Factor (HBEGF), Amphiregulin (AREG), Betacellulin (BTC), Epiregulin (EREG), Neuregulins (NRG1–4) and Epigen (EPGN). Ligands are normally expressed as transmembrane precursors and act in autocrine, paracrine and juxtacrine manner to influence EGFR activation. EGFR ligand precursors normally undergo ectodomain shedding by the action of membrane proteases of the ADAM family and activate the receptors (Edwards et al., 2008). The discovery of the “triple membrane passing signal” where a stimulus would trigger ADAM proteases activity and transactivate EGFR revealed multiple ways of regulating EGFR activity through external stimuli (Liebmann, 2011). Mouse models with aberrant expression of receptors or ligands show a profound effect on skin biology. For example, loss of TGF-alpha affects hair follicle structure while the overexpression of the same ligand or overexpression of the isoform ERBB2 causes skin hyperplasia and enhanced tumor formation. In contrast, deletion of EREG, or overexpression of AREG, cause inflammatory skin diseases with features of atopic dermatitis and psoriasis respectively (Schneider et al., 2008). Mice with an EGFR-dominant negative mutation targeted to the epidermis have curled whiskers and short hair that becomes progressively sparse (Murillas et al., 1995) while a defect in hair follicle differentiation and maturation is evident also in skin grafts from EGFR ablated neonatal skin on nude mice (Hansen et al., 1997).

Downstream effectors of EGFR signaling

After the transphosphorylation of the receptor on several tyrosine residues, a number of adapter proteins are recruited to the intracellular domain of EGFR. These include growth-factor-receptor bound-2 (GRB2) coupled to the guanine nucleotide releasing factor SOS. These protein interactions bring SOS in close proximity to Ras, allowing for Ras activation. Activated Ras recruits c-Raf to the cell membrane where it phosphorylates and activates MEK1 and MEK2, eventually leading to the activation of ERK1/2 (Oda et al., 2005). The EGFR/ERK pathway mediates both the pro-survival and proliferative programs of keratinocytes (Dumesic et al., 2009). ERK1/2 and JNK1/2 kinases are activated within the first few minutes of EGFR ligand binding and mediate expression of immediate early genes involved in keratinocytes proliferation. Within the early genes, c-jun is an important regulator of the EGFR pathway. c-Jun mutant primary keratinocytes exhibit a severe proliferation and differentiation defect and EGFR expression is downmodulated (Zenz et al., 2003). In the absence of fully functional EGFR, c-Jun expression and c-Jun dependent transcription are reduced in cultured keratinocytes and in skin (Mascia et al., 2010).

EGFR recruits directly through tyrosine phosphorylation phosphatidylinositol 3’-OH kinase (PI3K). The phospholipid products of PI3K activate phosphoinositide-dependent kinase 1 (PDK1) and recruits AKT to the plasma membrane. This sequence of events occurs upon UVB irradiation of keratinocytes and ROS mediated EGFR transactivation. Pretreatment of keratinocytes with EGFR inhibitors abolished UV-induced AKT activation-phosphorylation, and increases UV-induced apoptosis (Wang et al., 2003). Phospholipase Cγ1 associates with and is phosphorylated by the EGFR tyrosine kinase leading to mitogenic and differentiation responses involving PKC modification. For example, calcium induced TGF-alpha inhibits PKC-delta activity via EGFR dependent tyrosine phosphorylation, and this modification is associated with reduced keratinocytes terminal differentiation (Denning et al., 1996).

Deactivation of EGFR signaling

The multifunctional nature of EGFR activation requires multiple mechanisms of attenuating or eliminating the initial signal. Multiple negative feedback loops are placed in action including changes in compartmentalization, posttranslational modification, and transcriptional upregulation of negative regulators. Receptor internalization coupled to degradation is a highly effective process that attenuates EGFR signaling by targeting surface receptors for degradation in lysosomes (Sorkin and Goh, 2008). Another general mechanism of signal attenuation is represented by tyrosine phosphatases at the receptor level (Xu et al., 2005) and dual specificity phosphatases at the MAPK level (Zhang et al., 2010). A second wave of modulators becomes available upon transcriptional activation. For example, mitogen-inducible gene-6 (MIG6) is a newly synthesized negative regulator that functions directly on the receptor (Ferby et al., 2006).

EGFR is a major regulator of skin immunohomeostasis

While considerable work has focused on the role of the EGFR on epidermal and hair follicle development, keratinocyte proliferation and survival, recent studies have revealed that EGFR is a major contributor to immune homeostasis in the skin. Pharmacologic inhibition of EGFR regulates two subsets of cytokines and chemokines in keratinocytes at the transcriptional and posttranscriptional level. While the transcription of GM-CSF and CXCL8 is positively regulated with EGFR stimulation, absence of a functional EGFR/ERK pathway enhances expression of CCL2, CCL5 and CXCL10 and decreases expression of CXCL8 and GM-CSF both in vitro and in vivo (Mascia et al., 2010; Mascia et al., 2003; Pastore et al., 2005). EGFR blockade is also associated with increased levels of CCL17 and MHC class I and II (Komine et al., 2005; Pollack et al., 2011) in the skin and CXCL9, CCL3 and CCL11 in cervical cancer cell lines (Woodworth et al., 2005). The strongest evidence that EGFR signaling regulates inflammatory responses in human skin is provided by the adverse effects of the chemotherapeutic agents that target EGFR such as Cetuximab, Panitumumab, Gefitinib and Erlotinib. Within the first 2 weeks of treatment, the majority of cancer patients that receive anti-EGFR drugs develop a papulo-pustular rash associated with dry skin and pruritus (Lacouture, 2006). The presence of the rash is a general indicator that the drug hits its target both in the skin and in the tumor. Some authors speculate that there is an association between the rash grade and the overall survival of the patients (Perez-Soler and Saltz, 2005). Interestingly, the increase of homeostatic chemokines CCL27 and CXCL14 after treatment of cell lines with anti-EGFR drugs is associated with a possible role of these chemokines in the anti tumor immune response (Ozawa et al., 2009; Pivarcsi et al., 2007). These observations need to be corroborated in proper tumor models but underline the role of EGFR in regulating the expression of inflammatory mediators during skin carcinogenesis. Recent evidence shows that oncogenic Ras driven expression of CXCR2 ligands depends on EGFR expression and acts in an autocrine fashion to enhance migration of transformed keratinocytes (Cataisson et al., 2009). These data collectively define the profound influence of EGFR signaling on cutaneous biology.

The Multiple Contributions of Notch Signaling in Skin Biology

Overview of the Notch signaling pathway and relevance in skin biology

The Notch pathway constitutes a short-range communication channel involved in many fundamental aspects of multi-cellular life: proliferation, stem cells and niche maintenance, cell fate acquisition, differentiation and cell death. Mammals express multiple ligands and four Notch receptors but only one nuclear signal mediator, the DNA-binding protein RBPjk (Kopan and Ilagan, 2009)(Figure 4). Signaling is initiated when ligand binding to the receptor induces unfolding of a juxtamembrane negative control region, which allows access to proteases. Proteolysis by an ADAM sheds the extracellular domain, γ-secretase then cleaves Notch within its transmembrane domain to release the intracellular domain (Notch ICD, NICD or Notch intra). NICD translocates to the nucleus, binds to RBPjk and helps recruit one of of three adaptor proteins (MAML1–3; (Kopan and Ilagan, 2009)). Thus, every cleaved Notch molecule generates one signaling unit. In addition to this generalized model, evidence are mounting that RBPjk-independent activities of Notch can also contribute to vertebrate skin development (Demehri et al., 2008; Rangarajan et al., 2001), but the biochemical details of this signaling arm are yet to be uncovered.

Figure 4: Integration of Notch signaling circuitry:

The Notch pathway generates a short-range communication channel utilized throughout skin development and elsewhere in the adult to regulate multiple cellular processes (e.g., proliferation, stem cell and stem cell niche maintenance, cell fate specification, differentiation and cell death). Notch signals in a unique mechanism mediated by proteolysis and without any secondary messengers. Post-translational modifications and trafficking of the ligands and receptors can regulate the amplitude and timing of Notch activity. 1–2) Receptor Maturation. Upon translation, the Notch protein is fucosylated by the Pofut1. The fucose can be extended by the glycosyltransferase activity of Fringe. This impacts the ability of specific ligands to activate Notch (see below). During exocytosis the Notch receptor is cleaved by PC5, a protein convertase, at Site 1 (S1). At the cell surface Notch is a heterodimer (HD) held together by non-covalent interactions within the HD domain which, together with three Lin-Notch repeats (LNR) forms the negative regulatory region (NRR). 2) The steady-state levels of the Notch receptor at the cell surface are regulated by several proteins, many of which are E3 ubiquitin ligases (Dtx, Nedd4). After recycling, inappropriate receptor activation within endosomes can occur in the absence of ligand binding (NICD*). The ESCRT complex proteins are involved in Notch down-regulation, and mutations in this complex may contribute to pathogenesis in different cellular contexts. 3) Ligand maturation. Notch ligands are also Type I transmembrane proteins. The two major classes of ligands are Delta and Jagged (Serrate in Drosophila), the latter containing a cysteine rich domain. To signal, ligands must undergo ubiquitinylation by E3 ubiquitin ligases Neur and Mib which trigger Epsin-mediated endocytosis, and an undefined modification produces an active ligand which recycles to the cell surface in a Rab11 dependent process. Current models explaining the nature of ligand modification include ligand clustering, post-translational modifications and/or recycling into specific membrane domains. 4–6) Activation. Productive receptor-ligand interactions occur between neighboring cells (in TRANS) whereas negative interactions occur between receptor-ligand proteins coexpressed in the same cell (in CIS). Fringe-modified Notch1 receptors favor Delta binding (shown). Ligand endocytosis is thought to generate sufficient force to unfold the NRR, exposing Notch to cleavage at site S2 by ADAM10 metalloproteases. The membrane-anchored NEXT (Notch extracellular truncation) fragment is recognized by γ-secretase, an enzymatic complex composed of PS, NCT, PEN2 and APH1. γ-secretase then cleaves the Notch transmembrane domain sequentially starting near the cytosolic surface (sites S3 and S4) to release the Notch intracellular domain (NICD) and Nβ peptides, respectively. 5) In the absence of NICD, the DNA-binding protein CSL associates with ubiquitous co-repressor (Co-R) proteins, HDACs and Sirt1 to repress transcription of target genes. NICD binding to CSL recruits the adaptor protein Mastermind (MAM), which recruits the mediator complex and assembles an active transcription complex on target promoters. 6) During the transcriptional activation process, NICD is phosphorylated on its PEST domain by kinases such as CDK8 and targeted for proteasomal degradation by E3 ubiquitin ligases such as Sel10/Fbw7. This terminates the Notch signal and resets the cell for the next round of signaling. This figure is reproduced from (Ilagan and Kopan, 2007) with permission from the publisher.

Ignoring Notch function in the innate immune system, the Notch pathway participates in the development, maintenance or function of at least four skin compartments: the melanocytes (Kumano et al., 2008; Moriyama et al., 2006), the dermal papilla fibroblast (Hu et al., 2010), the epidermal keratinocyte (Blanpain et al., 2006; Demehri et al., 2008; Dumortier et al., 2010; Rangarajan et al., 2001), and the hair follicle keratinocyte (Lee et al., 2008; Moriyama et al., 2008; Pan et al., 2004). Cell autonomous Notch signals (that is, impacting the cell receiving the Notch signal) control melanocytes stem cells homeostasis, but Notch contributes to the melanocyte niche in the hair follicle also by regulating Kit ligand production (Lee et al., 2007) which is non cell autonomous (that is, by regulating diffusible or surface molecules affecting neighboring cells). Genetic loss of function analysis established that Notch does not contribute to the positional patterning of skin appendages, but once their position has been determined via Wnt, FGF, BMP and Edar signaling, Notch contributes in several ways to the normal formation of nails, hair follicles, feathers and scales. First, it is active in dermal papilla cells up stream of Wnt5A (Hu et al., 2010). Second, Notch receptors play important late role in maintenance of all the cell layers within the appendages. This function is both cell autonomous (Pan et al., 2004) and non-cell autonomous (Lin et al., 2000), including feedback interaction between the follicular keratinocyte, the dermal papilla and the melanocyte (Lee et al., 2007). Third, Notch-dependent signals either attract or support adipocytes as the dermis underneath Notch-deficient epidermis is depleted of lipid droplets and sebaceous glands form but do not mature (Pan et al., 2004). Fourth, in epidermal keratinocytes Notch signaling promotes exit from the basal layer and initiation of differentiation (Blanpain et al., 2006; Demehri et al., 2008; Dumortier et al., 2010; Rangarajan et al., 2001). Finally, Notch may contribute to the regulation of the stem cell niche in the human epidermis (Lowell et al., 2000).

Notch signaling in skin carcinogenesis

Loss of Notch1 in epidermis enhances carcinogenesis (Nicolas et al., 2003). This is due to a feed-forward, non cell autonomous loop between Notch-deficient keratinocytes, the dermal fibroblast population and the inflammatory infiltrate (Demehri et al., 2009). The resulting fibroplasia, inflammation and angiogenesis promote papilloma formation in both Notch-depleted and Notch-expressing keratinocytes within this environment. In this model progression to cancer is accelerated when Notch1 protein is lost (Demehri et al., 2009). Interestingly, skin keratinocytes are exquisitely sensitive to the dose of Notch signals. Spontaneous tumor latency decreases as the dose of Notch alleles decreases (Demehri et al., 2009). Mice lacking alleles of three out of four proteins comprising the γ-secretase complex (Pen2, Psen1, Ncstn) develop spontaneous tumors as well (Li et al., 2007; Xia et al., 2001). Importantly, humans with nonsense mutations in PEN2 (2 families), PSEN1 (1 family) and NCSTN (3 families) (Wang et al., 2010) develop hidradenitis suppurativa (HS) or acne inversa (AI; OMIM 11405761). Retroactive analysis of the Swedish national cancer registries identified a 50% increased risk for non-melanoma skin cancer in families with HS (Lapins et al., 2001). These findings dovetail nicely with recent reports that the γ-secretase inhibitor semagacestat, designed to reduce Notch signaling, was withdrawn from clinical testing due to elevated incidence of skin cancer. Given that heterozygosity of γ-secretase components is symptomatic in man and in mice, even “Notch-sparing” drugs must be carefully evaluated in clinical trials for skin differentiation defects and the subsequent inflammatory cascade.

Figure 3. Integration of EGFR signaling circuitry:

EGFR and other ERBB receptors dimerize after engagement with their ligands activating multiple signaling cascades through recruitment of adaptor and signaling proteins to the intracellular portion of the receptor. Phospholipase Cγ (PLC γ) activity triggers the activation or downmodulation of protein kinase C isoforms (PKCs) and increases intracellular calcium leading to the regulation and expression of proteins that control differentiation. Phosphoinositol 3’-OH kinase (PI3K) promotes survival responses through AKT activity. Activation of Ras regulates JNK1/2 and ERK1/2 MAPKs activity causing changes in gene transcription and mRNA stability. As a consequence, cell cycle progression and inflammatory responses are fine tuned by EGFR activity.