The pleiotropic roles of autophagy regulators in melanogenesis

Authors


Anand K. Ganesan, e-mail: aganesan@uci.edu

Summary

Melanin pigments protect the skin and eyes from toxic insults and are critical for the normal functioning of multiple organ systems including the skin, eyes, and brain. Biochemical and genetic studies in both human and mice have revealed the molecular machinery controlling the transcription of genes encoding enzymes that produce melanin and the trafficking of these enzymes to the melanosome, a lysosome-related organelle dedicated to melanin synthesis. Recent functional genomic studies have identified a role for genes previously known to regulate autophagy, a cellular process that facilitates nutrient recycling during starvation, in the biogenesis of melanosomes in vitro and in vivo. In this review, we describe the pleiotropic roles of autophagy regulators in multiple vesicle trafficking processes, define a specific role for autophagy regulators in melanosome biogenesis, and shed light on how autophagy and autophagy regulators may play different roles in both the biogenesis of melanosomes and melanosome destruction.

Introduction

Melanin, pigment produced from tyrosine that absorb light over both the UV and visible spectrum (Stein, 1955), protects the skin, eyes, and brain from toxic insults (Slominski et al., 2004) and also plays a critical role in the adaptation of organisms to their environment (Hoekstra, 2006). Defects in melanogenesis impact the physiology of multiple organ systems and contribute to the etiology of acquired human disorders. The skin pigmentary disorders vitiligo and melasma are characterized by aberrant regulation of melanogenesis (Grimes et al., 2006). For example, vitiligo melanocytes accumulate abnormal melanosome structures (Boissy et al., 1983, 1991a,b) and this disorder has been linked with polymorphisms in the tyrosinase gene (Jin et al., 2010), implicating a role for defects in melanogenesis in the pathogenesis of this disease. Similarly, age-related macular degeneration, the most common cause of blindness in the USA (Sarangarajan and Apte, 2005), is characterized by irregular deposition of melanin in the macula. Neuromelanin appears to play a dual role in the pathogenesis of Parkinson’s disease (Zecca et al., 2006), a disorder characterized by the loss of both dopaminergic neurons and melanin in the substantia nigra and locus ceruleus.

Genetic mutations in key pigment regulatory genes are known to result in monogenic human disorders and murine coat color defects (Bennett and Lamoreux, 2003). Albinism is characterized by mutation in one of four known pigment regulatory genes (tyrosinase, P gene, Tyrp1 gene, or the MATP gene) that can also present with nystagmus, blindness, and increased risk of skin cancer (Okulicz et al., 2003). Waardenburg’s syndrome is a monogenic disorder characterized by loss of melanin in the skin and hearing loss due to a mutation in one of six different genes (Price et al., 1998). Mutation of multiple genes that regulate melanosome formation and transport result in both coat color defects in mice (Huizing et al., 2002) and the human pigmentary disorder Hermansky–Pudlak syndrome (Shotelersuk and Gahl, 1998).

Biochemical and genetic analyses of mice with coat color defects have led to the elucidation of complex cellular mechanisms that regulate melanogenesis (Bennett and Lamoreux, 2003). As the production of melanin is both metabolically expensive and generates free radicals (Meyskens et al., 2001), the melanocyte has evolved multiple transcriptional, enzymatic, and spatial control mechanisms to control melanogenesis (Yamaguchi et al., 2007). Melanin is synthesized within the melanosome, a lysosome-related organelle whose biogenesis is regulated by specific transport pathways and proceeds through four stages (Raposo and Marks, 2007). Stage I melanosomes, vesicles derived from early endosomal membranes (Raposo et al., 2001), contain the amyloid protein PMEL17 (Theos et al., 2005) and MART-1 (Hoashi et al., 2005). As the stage I melanosome matures, PMEL17 forms lumenal fibrillar striations that characterize stage II melanosomes (Hurbain et al., 2008) in a process that requires proteases (Berson et al., 2003). The resulting premelanosomes mature to stage III and IV melanosomes after the delivery of melanogenic proteins from early endosomes via vesicular transport (Gautam et al., 2006). Once these proteins are transported, melanin pigment is synthesized and deposited onto the Pmel17 striations (stage III melanosome), eventually giving rise to the stage IV melanosome, which appears opaque on electron microscopy (Hearing, 2005; Wasmeier et al., 2008; Yamaguchi and Hearing, 2009). Melanocyte-specific microphthalmia-associated transcription factor (MITF-M) is known to be the master regulator for the transcription of enzymes and proteins that are core components of the melanosome (Levy et al., 2006): TYR (the rate-limiting step in melanin synthesis), TYRP1, PMEL17, and MART1 (Cheli et al., 2010). The expression of MITF is controlled by at least five transcription factors, LEF1, CREBP, SOX10, OC2 and PAX3, which directly bind to the MITF-M core promoter (Steingrimsson et al., 2004). Multiple different intracellular signaling pathways further regulate the activity of these transcription factors (Costin and Hearing, 2007). In summary, mammalian cells have evolved complex mechanisms to ensure that melanin is only synthesized in the melanosome by precisely regulating the production and delivery of enzymes to this organelle.

Although many genes that regulate melanin production have been identified (Yamaguchi and Hearing, 2009), genetic studies suggest that many other genes influencing human pigment variation have yet to be elucidated (Sturm, 2009). A genome-wide RNA interference (RNAi) screen to identify novel regulators of melanogenesis identified 92 novel genes as putative regulators of melanogenesis, indicating that multiple genes and pathways may play incremental roles in regulating melanin production (Ganesan et al., 2008). Validation studies revealed that many of the novel regulators of melanogenesis control the expression of MITF and tyrosinase (Ganesan et al., 2008).Additionally, a number of putative regulators of autophagy, including genes that regulate the initial stages of melanosome formation (BECN1) (Funderburk et al., 2010) and genes that regulate the turnover of autophagic vesicles (WIPI1) (Obara and Ohsumi, 2008), were identified as potent regulators of melanogenesis (Ganesan et al., 2008). Immunohistochemical studies have determined that WIPI1 (Proikas-Cezanne et al., 2004), BECN1, and LC3 are expressed in melanocytic lesions (Miracco et al., 2010). Colocalization studies revealed that the autophagosome marker LC3 (Klionsky et al., 2008) is present in the melanosome (Ganesan et al., 2008), suggesting that regulators of autophagy may play a role in generating the melanosome. Consistent with this hypothesis, transgenic mouse studies revealed that mice deficient in genes that regulate autophagosome formation (BECN1) or genes that putatively control autophagosome turnover (Fig 4, Vac14) are associated with coat color defects in mice (Chow et al., 2007; Ganesan et al., 2008; Jin et al., 2008). The molecular target of rapamycin (mTOR) is a critical kinase involved in energy sensing, and inhibition of mTOR leads to the activation of autophagy (Meijer and Codogno, 2011). Patients with tuberous sclerosis have mutations in one of two genes (TSC1, TSC2) that act to suppress mTOR, leading to increased intracellular mTOR activity (Inoki et al., 2005; Kwiatkowski, 2003). One of the hallmarks of tuberous sclerosis is the development of hypopigmented macules early in life, which are characterized by decreased numbers of melanosomes (Jimbow et al., 1975), implicating a role for autophagy regulators in melanogenesis in human skin. Finally, pharmacologic inhibition of autophagy has potential therapeutic implications, as compound screens have already revealed that agents known to inhibit autophagosome turnover inhibit pigment production in skin equivalents (Ni-Komatsu et al., 2008). Taken together, these studies provide compelling evidence that autophagy is involved in the synthesis of melanin.

In this review, we highlight recent evidence that autophagy regulators control a broad range of cellular trafficking pathways and describe the specific contributions of autophagy regulators to both melanosome biogenesis and the transcription of MITF target genes. We also present the hypothesis that autophagy may control both the formation of the melanosome and the destruction of abnormal melanosomes in the context of human disease (Hermansky–Pudlak syndrome and melanoma).

Autophagy – a pleiotropic regulator of multiple vesicle trafficking pathways

Macroautophagy (referred to as autophagy in this review) is a highly conserved cellular pathway involved in tissue homeostasis, adaptation to starvation, and removal of dysfunctional organelles or pathogens (Levine and Kroemer, 2008). Dysfunction in autophagy has been associated with cancer formation, neurodegeneration, and aging (Huang and Klionsky, 2007; Levine and Kroemer, 2008; Mizushima et al., 2008).Detailed studies in both yeast and mammalian cells have determined that starvation-induced autophagy proceeds through three distinct steps: (i) vesicle nucleation, (ii) expansion and closure of the vesicle membrane, and (iii) fusion and recycling of autophagy protein complexes (Xie and Klionsky, 2007). Recent studies have uncovered molecular events that lead to vesicle nucleation. Starvation induces the activity of the small G-protein RalB (Bodemann et al., 2011). RalB activation promotes the formation of an autophagy complex that contains the exocyst complex (He and Guo, 2009) and the exocyst component Exo84, but not mTOR (Bodemann et al., 2011). mTOR normally acts to repress autophagy by phosphorylating and inactivating the ULK1-ATG13-FIP200 autophagy induction complex (Ganley et al., 2009; Jung et al., 2009). Once the ULK1-ATG13-FIP200 complex is active, the ULK1 kinase autophosphorylates, which allows the ULK1-ATG13-FIP200 complex to associate with ATG101 (Mizushima, 2010). This complex then leads to the activation of an autophagy-specific class III PI(3)K complex, the Beclin1-ATG14L-VPS34-VPS15 complex (Funderburk et al., 2010). This complex is critical for vesicle nucleation, as this complex can coat the cup-shaped isolation membrane with phosphatidylinositol-3-phosphate (PI(3)P), which then recruits ATG16, ATG5, and ATG12 to the membrane (Suzuki et al., 2001).

During the next phase of autophagy, expansion and closure of the vesicle membrane, ATG5 gets conjugated to ATG12 and LC3 gets conjugated to phosphatidylethanolamine via a process involving ATG3, ATG4, and ATG7 (Tanida, 2011; Tanida et al., 2004). Phosphatidylethanolamine-conjugated LC3 (LC3-II) then coats the inner and outer surface of the autophagosomes, and it has been identified as a marker of mature autophagosomes (Tanida et al., 2004). Prior to fusion with the lysosome, the autophagosome fuses with multivesicular bodies, an endosomally derived compartment, to form the amphisome (Fader and Colombo, 2009). The amphisome then ultimately fuses with the lysosome, which can then hydrolyze engulfed material (Fader and Colombo, 2009). Many of the protein complexes involved in autophagy are recycled so they can be used to generate additional vesicles (Yorimitsu and Klionsky, 2005). Proteins that bind to phosphatidylinositol, including the yeast protein ATG18 and its human homolog WIPI1, bind to the autophagosome-lysosome membrane and function to recycle membrane and autophagy protein complexes (Yorimitsu and Klionsky, 2005).Recent studies have uncovered evidence that in addition to regulating cellular self-degradation, the autophagy machinery plays a much broader role in regulating specific vesicle transport processes (Razi et al., 2009; Zhong et al., 2009). In yeast, two different pathways utilize the same autophagy machinery – the autophagy pathway involved in self degradation and the cytoplasm to vacuole targeting pathway, a specific hormone-responsive transport pathway that is active under nutrient-rich conditions (Xie and Klionsky, 2007). In mammalian cells, multiple different pathways utilize components of the autophagy machinery for vesicle transport. Beclin-1, a critical regulator of autophagy, associates with many different proteins that regulate diverse cellular processes from autophagy to apoptosis (Kang et al., 2011). Beclin-1 associates with two other proteins, VPS15 and VPS34, to form a core PI3-kinase complex that regulates vesicle trafficking (Figure 1) (Kang et al., 2011). This PI-3 kinase complex associates with additional proteins to form multiprotein complexes that regulate autophagy, endosome/autophagosome maturation, and inhibit autophagy, respectively (Zhong et al., 2009) (Figure 1). Under conditions of starvation, RalB activation promotes the assembly of a complex containing the exocyst complex, the exocyst component Exo84, the PI-3 kinase complex, and Atg14L, leading to the induction of autophagy (Bodemann et al., 2011). In the absence of starvation, the PI3 kinase complex can associate with the exocyst complex, the exocyst component Sec5, Rubicon, and mTOR, which leads to the repression of starvation-induced autophagy (Bodemann et al., 2011). This Sec5-PI3 kinase complex is localized to a different subcellular compartment than the Exo84-PI3 kinase complex (Bodemann et al., 2011). The Sec5-containing multiprotein complex is either an inactive autophagy complex as proposed (Bodemann et al., 2011) or a complex that regulates other vesicle transport processes, a contention which is supported by the documented role of Sec5 in the transport of membrane proteins in neuronal cells (Murthy et al., 2003).The PI3 kinase complex can also associate with UVRAG in a complex that does not contain ATG14L but does contain Rubicon (Matsunaga et al., 2009). This complex does not associate with the exocyst machinery (Liang et al., 2008). In normal cells, overexpression of UVRAG enhances both autophagosome maturation and the endocytic trafficking pathway, demonstrating that the beclin-1 UVRAG complex may play pleiotropic roles in both autophagy and normal endocytic trafficking (Liang et al., 2008).Similarly, molecular regulators of autophagy protein complex recycling are required for multiple vesicle transport processes. While the yeast protein ATG18 is involved in the recycling of membrane containing PI3P after fusion with the lysosome (Levine and Kroemer, 2008; Reggiori et al., 2004), it also acts as a phosphatidylinositol (PtdIns) 3,5P2 effector to remodel the membrane of vacuoles (Dove et al., 2004; Guan et al., 2001). Similarly, the mammalian homolog of ATG18 (WIPI1) localizes to multiple, different vesicle compartments under normal nutritional conditions (Jeffries et al., 2004), and localizes to the autophagosome under conditions of starvation or TORC1 inhibition (Proikas-Cezanne and Pfisterer, 2009). In COS-7 cells, WIPI-1 was shown to regulate trans-Golgi-endosomal protein trafficking (Jeffries et al., 2004), defining a role for WIPI1 in endocytic trafficking. WIPI1 also controls melanosome biogenesis, another process involving endocytic transport (Ho et al., 2011). In yeast, Atg18 (WIPI1 homolog) is known to associate with Fab1, Fig 4, Vac14, and Vac7, forming a PtdIns(3,5)P2-synthesizing complex in the yeast vacuole membrane (Efe et al., 2007). Consistent with a role for autophagy complex recycling in endosomal transport, the human Fab1 homolog phosphatidylinositol 3-phosphate 5-kinase (PIKfyve) has been shown to regulate trans-Golgi network to endosome retrograde transport (Rutherford et al., 2006). The role of this complex in autophagy protein complex recycling is depicted in Figure 1.

Figure 1.

 Multiprotein complexes differentially regulate intracellular vesicle traffic. Initiation of vesicle formation is regulated by one of three different macromolecular complexes: an ‘autophagy-on’ complex activated by nutrient deprivation (A), an ‘autophagy-off’ complex (B), and a complex of proteins that putatively regulates endosomal trafficking (C). These initiation complexes are composed of individual smaller protein complexes: the ULK1 complex (ULK1-ATG13-FIP200) and the PI3K complex (BECN1-VPS34-VPS15). Each of these large complexes can initiate the formation of three types of intracellular vesicles: autophagosomes (green membrane), melanosomes (black membrane), or endosomes (red membrane). Subsequent to vesicle formation, a complex containing WIPI1 and putatively containing Fig 4, and Vac14 catalyzes the turnover/remodeling of each of these vesicles. The generation of these two vesicles requires the UBL complex (ATG3, ATG4, ATG7, and the ATG5/12 conjugate) that catalyzes LC3 lipidation. As nutrient deprivation does not result in increased melanogenesis, melanosome biogenesis is regulated by either (B) or (C) (dashed arrows).

Transgenic mouse studies have revealed that mice deficient in components of this recycling complex possess defects in autophagy (Ferguson et al., 2010). Fig 4 mutant mice accumulate enlarged vesicles in the brain that stain with the autophagy markers p62 and the lysosome marker LAMP2 (Ferguson et al., 2009). Similarly, mice with mutations in Vac14 accumulate enlarged vacuoles in the brain containing GFAP (Ferguson et al., 2009). Fig 4 mutant mice have decreased accumulation of melanosomes in the mouse hair follicle (Chow et al., 2007). Vac14 also exhibits coat color defects, although the mechanism has not been characterized (Jin et al., 2008). These studies suggest that regulators of autophagosome turnover are also involved in melanosome biogenesis (Ferguson et al., 2010), a process that is controlled by endosomal transport. The findings are consistent with a model in which the machinery that regulates the formation of the autophagosome and the recycling of autophagy complexes play pleiotropic roles in other vesicle trafficking processes (Figure 1).

While it is clear that some autophagy regulators play pleiotropic roles in vesicle trafficking processes other than autophagy, mounting evidence also suggests that selective forms of autophagy exist which regulate key aspects of cell physiology. Recent studies in yeast have determined that starvation induces the secretion of acyl-coenzyme-A binding protein via a mechanism involving the autophagy machinery termed exophagy (Abrahamsen and Stenmark, 2010). Additionally, mice with mutations in the genes Atg4b and Atg5 have defects in the development of otoconia, an organelle in the inner ear that regulates balance, demonstrating a role for autophagy in the generation of this organelle (Marino et al., 2010). These findings demonstrate that both autophagy and autophagy regulators play different roles in multiple vesicle trafficking processes.

Melanosome biogenesis – an endosomal transport process that utilizes autophagy regulators

Accumulating evidence suggests that autophagy regulators play a key role in the formation of the melanosome. Genome-wide RNAi approaches identified three autophagy regulators as novel regulators of melanogenesis – MAP1LC3C, WIPI1, and GPSM1. Follow-up studies revealed that depletion of other autophagy genes (MAP1LC3A, BECN1) resulted in decreased melanin accumulation, and immunofluorescence studies revealed that autophagosome markers and melanosome markers colocalize (Ganesan et al., 2008). Follow-up studies demonstrated that chronic suppression of WIPI1 expression resulted in a decreased number of melanosomes, suggesting that these autophagy genes function to regulate melanosome formation (Ho et al., 2011). In addition to identifying an unappreciated link between autophagy and melanosome biogenesis, these findings raised a critical question: Does autophagy itself regulate the biogenesis of lysosome-related organelles or do regulators of autophagy play pleiotropic roles in the endosomal-based pathways that generate the melanosome? If autophagy regulates melanosome biogenesis directly, stimulation of autophagy via nutrient deprivation should result in melanin accumulation. Although starvation is sufficient to induce autophagosome formation, it is not sufficient to induce melanin accumulation, indicating that autophagy itself does not regulate melanogenesis (Ho et al., 2011). RNAi-based loss of function studies revealed that depletion of several key autophagy regulators resulted in changes in melanin accumulation, most notably WIPI1, BECN1, and mTOR. Taken together, these studies suggest that only select components of the autophagy pathway, namely regulators of vesicle formation and proteins that are involved in recycling components of the autophagosome membrane, are required for melanosome biogenesis (Ho et al., 2011).

One of the regulators of autophagosome turnover with potent impacts on melanin accumulation was WIPI1. WIPI1, a human homolog of the yeast protein Atg18, is composed of multiple WD40 repeat domains which allow it to bind to phosphatidylinositol 3-phosphate (PtdIns3P) and PtdIns3,5P2 (Jeffries et al., 2004; Proikas-Cezanne et al., 2007). While published studies (Jeffries et al., 2004; Proikas-Cezanne et al., 2007) suggest that WIPI1 functions primarily to control vesicle trafficking, WIPI1 depletion also significantly inhibited the accumulation of MITF and tyrosinase mRNA in MNT-1 cells (Ganesan et al., 2008) and normal melanocytes (Ho et al., 2011), even though WIPI1 has no DNA-binding domain or nuclear-localization signal (Jeffries et al., 2004). These studies implicate a role for regulators of autophagy protein complex recycling in both regulating the transcription of melanosome cargo and in potentially regulating the formation of the melanosome itself.

Detailed studies sought to identify autophagy regulators that control the transcription of MITF and MITF-M target genes. Depletion of mTOR stimulated the accumulation of MITF mRNA, whereas depletion of WIPI1 inhibited the accumulation of MITF mRNA (Ho et al., 2011). Overexpression and RNAi approaches revealed that WIPI1 regulates the transcription of MITF and MITF target genes. In contrast, depletion of mTOR or pharmacologic inhibition of TORC1 led to increased MITF transcriptional activity. Additional studies revealed that depletion of WIPI1 increased TORC1 activity, leading to the repression of TORC2, activation of GSK3β, increased β-catenin degradation, and decreased MITF transcription (Ho et al., 2011). As β-catenin is a central regulator of MITF (Larue et al., 2003), these studies indicate that WIPI1 modulates mTOR activity, leading to changes in β-catenin stability and are consistent with other studies demonstrating that rapamycin treatment stimulated melanogenic enzyme transcription (Busca et al., 1996; Ohguchi et al., 2005) and GSK3β inhibition led to increased melanogenesis (Bellei et al., 2008). The studies indicate that WIPI1 modulates MITF and tyrosinase transcription by modulating mTOR activity (Ho et al., 2011).

In addition to regulating the transcription of genes encoding enzymes that are transported to the melanosome, both WIPI1 and mTOR regulate the formation of the melanosome itself. WIPI1-deficient melanocytic cells accumulated small (0.4 μm) single-membrane vesicles with intracellular vesicles within them resembling stage I melanosomes (Ho et al., 2011). WIPI1-depleted cells also accumulated the autophagy markers p62 and LC3-II (Ho et al., 2011). As the stage I melanosome marker Pmel17 colocalizes with p62 and LC3-II (Ganesan et al., 2008), these studies suggest that the stage I melanosome contains both p62 and LC3-II. In contrast, mTOR inhibition resulted in increased accumulation of mature melanosomes, decreased accumulation of p62, and increased accumulation of LC3-II, consistent with data that p62 colocalizes with the early but not late melanosomes (Ho et al., 2011).

Becn1 depletion inhibited pigment accumulation in vitro and in vivo (Ho et al., 2011) and inhibited the accumulation of LC3-II and melanosomes (Ho & Ganesan, unpublished observation), while not affecting the accumulation of MITF and tyrosinase mRNA (Ho et al., 2011).

These observations suggest that autophagy regulators control the formation and maturation of the stage I–II melanosome, a vesicle that is endosomally derived. As starvation does not enhance melanosome biogenesis and the stage I–II melanosome is derived from the lysosome and not the endosome, the regulation of melanosome formation by autophagy regulators is likely independent of the well characterized autophagosome–lysosome nutrient recycling pathway.

Other studies have also suggested that autophagosomes or autophagosome-like intermediates may have a role in melanosome biogenesis, a process that is controlled by endosomal trafficking. Disruption of ESCRT-I in melanocytic cells resulted in the accumulation of both Tyrp1 and MART-1 in autophagosome-like structures that were derived from endosomal membranes, suggesting that these structures could be intermediates in the trafficking of these proteins (Truschel et al., 2009). Pharmacologic inhibitors of autophagosome turnover also inhibited pigment accumulation (Ni-Komatsu et al., 2008). Finally, mice defective in autophagy complex recycling (Fig 4 mutant mice) or humans with mutations that inhibit autophagosome formation (tuberous sclerosis) exhibited decreased melanosome accumulation (Chow et al., 2007; Jimbow et al., 1975). These studies strongly support the hypothesis that autophagy regulates melanosome accumulation.

Whereas some observations suggest that autophagy regulators may play a primary role in regulating the trafficking and maturation of the melanosome, other observations suggest that autophagy regulators function mainly to control the transcription of melanogenic enzymes via a mechanism involving mTOR (Figure 2A). MITF is a known regulator of many components of the early stage melanosome, including both Pmel17 and MART-1 (Du et al., 2003). Depletion of WIPI1 in primary melanocytes resulted in decreased accumulation of Pmel17 but not MART-1, indicating that the decreased accumulation of mature melanosomes observed could be via a primary impact on MITF transcription. Consistent with this hypothesis, mTOR depletion resulted in the increased accumulation of MITF and an increased accumulation of Pmel17.

Figure 2.

 Autophagy regulators control melanosome biogenesis and the transcription of melanogenic enzymes. (A) Autophagy regulators control MITF transcription which then regulates melanosome biogenesis. WIPI1 modulates molecular target of rapamycin activity which leads to β-catenin stabilization and increased MITF transcription. MITF regulates proteins involved in melanosome biogenesis and the transcription of terminal melanosome enzymes. (B) Autophagy regulators control melanosome biogenesis with secondary effects on transcription. Melanosome formation is likely regulated by either the autophagy-off complex (b) or the endosomal trafficking complex (c). Once the early stage melanosome is formed, WIPI1 and proteins that interact with WIPI1 (Fig 4, Vac14) interact with the stage I–II melanosome and regulate its maturation. WIPI1 and proteins complexed to WIPI1 can then modulate TORC1 signaling, leading to increased β-catenin nuclear localization and increased MITF transcription. MITF regulates the transcription of Tyr and Tyrp1, which are then transported to the stage III melanosome. During melanosome maturation, p62 is degraded while LC3 continues to remain on the melanosome membrane as indicated.

Becn1 depletion had no impact on the transcription of melanogenic enzymes while inhibiting LC3-II and melanosome accumulation (Ho & Ganesan, unpublished observation), indicating that not all autophagy regulators control MITF and tyrosinase transcription. As Becn1 controls melanosome formation but not transcription, it is conceivable that autophagy regulators may function primarily by controlling vesicle trafficking with secondary effects on transcription, consistent with the known function of these genes.

Future studies will be required to determine whether autophagy regulators control melanogenesis directly by controlling MITF transcription or indirectly by controlling vesicle trafficking which can then feedback to regulate transcription.

Although these studies have identified individual autophagy genes that regulate melanogenesis, it is less clear what protein complexes are involved in this process. Starvation induces the redistribution of proteins between an ‘autophagy-off’ complex (Sec5 complex) and an ‘autophagy-on’ complex (Exo84) (Bodemann et al., 2011). Whereas chronic suppression of mTOR via starvation was sufficient to promote the formation of the Exo84 complex (Bodemann et al., 2011), transient mTOR inhibition was unable to induce the disassembly of the Sec5 complex (Bodemann et al., 2011), suggesting that transient inhibition and chronic inhibition of mTOR have differential effects on autophagy induction.

In melanocytic cells, transient inhibition of mTOR activity was sufficient to induce melanosome maturation, but starvation was not (Ho et al., 2011), indicating that the Exo84 complex does not regulate melanosome biogenesis. Melanosome biogenesis could be regulated by either the Sec5 complex or the UVRAG complex, and polymorphisms in the UVRAG gene have been linked to vitiligo (Jeong et al., 2010).

Future studies will determine which PI3K complex regulates the formation of the stage I–II melanosome whether WIPI1 is a component of these complexes or modulates their activity, and how these complexes may modulate MITF transcription.

Autophagy regulators in melanosome formation and melanosome destruction

In addition to its recently identified role in melanogenesis (Ho et al., 2011; Ganesan et al., 2008), autophagy also plays an established role in the removal of defective cellular organelles in specific cell types. For example, zymophagy is a recently characterized selective form of autophagy that regulates the removal of pancreatic enzymes released during pancreatitis (Grasso et al., 2011). Pexophagy is a selective form of autophagy that controls the degradation of defective peroxisomes (Sakai et al., 2006). Similarly, autophagy likely plays a role in the selective removal of defective melanosomes.

Autophagosomes accumulate in cells obtained from patients with HPS-1, a disorder of melanosome biogenesis, suggesting that their presence is secondary to an attempt to remove defective melanosomes (Smith et al., 2005). Monobenzone induces T-cells to attack the melanocyte, resulting in the engulfment of melanosomes within autophagosomes (Van Den Boorn et al., 2011). Autoimmune chicken models of vitiligo also exhibit the accumulation of abnormal melanosomes (Boissy et al., 1986). Finally, subsets of both human (Lazova and Pawelek, 2009; Lazova et al., 2010) and rodent (Bomirski et al., 1987) melanomas are characterized by the accumulation of autophagosomes which often contain preformed melanosomes (Bomirski et al., 1987), defining a role for autophagosomes in degrading the melanosome in melanoma.

Recent studies have indicated that different protein complexes are involved in the formation of autophagosomes and endosomes. Starvation induces the relocalization of autophagy regulators from an ‘autophagy-off’ complex containing Sec5 to an ‘autophagy-on’ complex containing ATG14L. Another complex composed of the PI3 kinase complex and UVRAG is known to regulate endosomal trafficking. mTOR inhibition in melanocytic cells induces the formation of melanosomes but not autophagosomes, suggesting that melanosome formation does not involve the ‘autophagy-on’ complex (Figure 3). As melanosome destruction involves the formation of true autophagosomes, a process regulated by the ATG 14L complex (Figure 3). Future studies will delineate the differential roles of multiprotein complexes composed of autophagy regulators in melanosome formation and melanosome destruction.

Figure 3.

 Autophagy regulators control melanosome formation and melanosome destruction. Melanosome formation is initiated by either the autophagy-off complex (B) or the endosomal trafficking complex (C). This results in the formation of the early stage melanosome, whose turnover is regulated by the WIPI1-Fig 4-Vac14 complex. Under conditions of stress, starvation, or defective melanosome production, the ‘autophagy-on’ (A) complex can be activated, resulting in the formation of the autophagosome, which then engulfs and degrades the melanosome.

Concluding remarks and perspective

Recent studies have determined that autophagy regulators control multiple different vesicle trafficking pathways, ranging from biosynthetic pathways (melanogenesis), to pathways involved in the degradation of organelles (pexophagy), to pathways involved in the recycling of nutrients during starvation. More importantly, autophagy regulators play a key role in regulating melanocyte biology – they play an intimate role in controlling both the production of the melanosome and the transcription of MITF target genes. In addition to regulating melanosome formation, conventional autophagy plays a role in removing melanosomes in the context of human disease and melanoma. Recent studies have provided the insight that different multiprotein complexes play roles in regulating autophagy and endosomal trafficking. Future studies will characterize the differential role of multiprotein autophagy complexes in melanosome formation and melanosome destruction.

Acknowledgements

This work was supported by grants to Anand Ganesan from National Institute of Health (1RO3AR057150 and 1K08AR056001) and a Dermatology Foundation research grant. Hsiang Ho was supported by a Joanna M. Nicolay Melanoma Foundation Research Scholar Award and a Graduate Fellowship in Alternatives in Scientific Research.

Ancillary