COMMENTARY
Tarafder AK, Bolasco G, Correia MS et al. (2014) Rab11b mediates melanin transfer between donor melanocytes and acceptor keratinocytes via coupled exo/endocytosis. J Invest Dermatol 135:1056–66 Van Den Bossche K, Naeyaert JM et al. (2006) The quest for the mechanism of melanin transfer. Traffic 7:769–78 Wasmeier C, Romao M, Plowright L et al. (2006) Rab38 and Rab32 control post-Golgi trafficking of melanogenic enzymes. J Cell Biol 175:271–81
Weiner L, Han R, Scicchitano BM et al. (2007) Dedicated epithelial recipient cells determine pigmentation patterns. Cell 130: 932–42 Wu XS, Masedunskas A, Weigert R et al. (2012) Melanoregulin regulates a shedding mechanism that drives melanosome transfer from melanocytes to keratinocytes. Proc Natl Acad Sci USA 109:E2101–9 Yamaguchi Y, Hearing VJ (2009) Physiological factors that regulate skin pigmentation. Biofactors 35:193–9
See related article on pg 1067
Hiding in Plain Sight: Molecular Genetics Applied to Giant Congenital Melanocytic Nevi Heather C. Etchevers1 Large and giant congenital melanocytic nevi are rare malformations that offer surprising insight into prenatal and postnatal acquisition of nevi of any size, central and peripheral nervous system development, and melanomagenesis. In this issue, Charbel et al. demonstrate the use of highly sensitive detection techniques for recurrent but difficult-to-detect mutations in NRAS and BRAF. It is now possible to systematically add a molecular qualifier to distinguish lesions that had once been considered to be equivalent based on the single visual parameter of size. These findings help to elucidate the pathophysiology of congenital melanocytic nevi and their predisposition to malignancy. Journal of Investigative Dermatology (2014) 134, 879–882; doi:10.1038/jid.2013.531
Nosology of large and giant congenital melanocytic nevi
Congenital melanocytic nevi (CMN) are malformations resulting from the faulty development of melanocyte progenitors in the embryo or fetus, and they are composed ultimately of an abnormal mixture of skin cellular elements. These pigmented hamartomas occupy sharply defined areas along the epidermal– dermal junction that range from a few millimeters in diameter to large swathes of the body, limbs or head. In the larger forms, CMN (single or multiple) often extend vertically into the deeper dermis and more rarely into the hypodermis or even subcutaneous tissues. The first descriptions of children with large CMN date from observations recorded by the French Count of Buffon in 1777, but the
incidence of CMN seems to be independent of ethnic factors. CMN have been classified historically according to their predicted largest diameter in adulthood, as if they were circular (predicted adult size (PAS), as they enlarge proportionately to the child’s growth). Until recently, other qualifiers had not systematically been taken into account. CMN measuring 40 cm or larger in PAS are referred to as ‘‘giant’’ nevi, according to the most recent and complete classification system (Krengel et al., 2013). This is intended to promote an international standard in phenotyping by including additional criteria such as pilosity, color heterogeneity, rugosity, and the presence of nodular growths, with charts for predicting PAS according
1
INSERM, UMR_S910, Aix-Marseille Universite´, Marseille, France
Correspondence: Heather C. Etchevers, INSERM, UMR_S910, Aix-Marseille Universite´, Faculte´ de Me´decine, 27 Boulevard Jean Moulin, Marseille 13005, France. E-mail: [email protected]
to body site and photographic examples. Small (o1.5 cm PAS) CMN occur in more than 1 in 100 births. Larger CMN (Figure 1a) form a much rarer subset, with prevalence estimated at around 1 in 20,000 to 50,000 births as a function of size. Treatment options in the year 2014 remein exclusively surgical. The differential diagnosis of small and medium CMN includes smooth muscle hamartoma or Becker’s nevus, mastocytoma, variants of dermal melanocytosis, and cafe´-au-lait macules. Larger CMN have been confused with pigmented plexiform neurofibromas. Histologic evaluation, dermoscopic evaluation, and the development of typical CMN features over time may clarify the diagnosis. However, molecular characterization promises to be an important additional criterion in phenotyping even the most easily diagnosed lesions and in teasing prognostic factors for syndromic features such as symptomatic neurocutaneous melanocytosis or malignant melanoma. Sensitive molecular diagnoses, such as those most recently developed in this issue of JID by Charbel et al. (2014), may soon augment the diagnostic and therapeutic arsenal. ‘‘Satellite’’ is a term used commonly to describe small or medium CMN, or those that appear postnatally (‘‘tardive’’ nevi) in the presence of a large/giant CMN. Both semantically and molecularly, it is more accurate to refer to ‘‘disseminated’’ nevi (Kinsler et al., 2013). These disseminated CMN may be present at birth and/or may increase to significant numbers over the first few years of life. Occasionally, a single largest CMN cannot be considered to be the principal malformation among the many, in which case the child is said to have ‘‘multiple medium’’ CMN. This presentation correlates with an increased predisposition to neurological abnormalities, as are greater satellite numbers and PAS420 cm (large and giant CMN; Shah, 2010). Contrary to a common interpretation of the earliest reports of such correlations, there is no discrete biological cut-off as to how many additional disseminated CMN predispose to neurological symptoms or melanoma. www.jidonline.org
879
COMMENTARY
Clinical Implications
Gain-of-function somatic mutations in NRAS at Q61 have been found in large and giant congenital melanocytic nevi (CMN), but mutant alleles can be difficult to detect.
Gain-of-function somatic mutations in either BRAF at V600 or NRAS Q61 have been found in small and medium CMN.
The principal CMN harbors the same mutation as found in samples from disseminated nevi or complications such as melanoma or central nervous system anomalies of the patient, implying their common origin.
within the malformation, but it is often extracutaneous, with some predilection for the CNS. Other malignancies such as liposarcomas, rhabdomyosarcomas, and peripheral nerve sheath tumors have also been described in association with large CMN. Rapidly growing ‘proliferative nodules’ within the CMN can mimic melanoma but often show benign features on closer examination (Phadke et al., 2011). Molecular basis of nevus formation
Syndromic CMN
Neurocutaneous melanocytosis, cited in earlier literature as neurocutaneous melanosis (NCM), is characterized by abnormal aggregations of nevomelanocytes within the central nervous system (CNS) of an estimated 5–15% of patients with the larger forms of CMN, or multiple medium-sized CMN. Neurological signs may include hydrocephalus, epilepsy, arachnoid cysts, tethered spinal cord, Dandy–Walker malformation (cerebellar vermis hypoplasia), developmental delay, and a number of rarer CNS tumors (Kinsler et al., 2013). True melanosis, through accumulation of
Both CMN and acquired melanocytic nevi are associated exclusively with somatic mutations in intracellular proteins of the microtubule-associated protein kinase (MAPK) signal transduction pathway (Figure 1b). In some series of CMN, only a recurrent mutation in the NRAS gene at codon 61 has been described, transforming the glutamine (Q) to lysine (K) or arginine (R) (Kinsler et al., 2013, and references therein), but many others have described a single hotspot mutation in BRAF (V600E) as well (Dessars et al., 2009; Phadke et al., 2011)—as is found commonly in adultonset melanoma. All of the identified
melanin in CNS neurons or glia, may also occur in association with CMN, but further neuropathological studies need to be undertaken for this to be conclusive. Melanoma develops in an estimated 1–2% of pre-pubertal patients with large CMN or NCM, on the order of 10,000 times more often than the incidence among the general population of children under the age of 10, based on registry data from the SEER (Surveillance, Epidemiology, and End Results) program of the US National Cancer Institute, 1975–2000. CMN-associated melanoma may be cutaneous, usually
a
b DKK1 WNT
FZ
KITL
FGF
EGF
KIT
FGFR
EGFR
HGF
MET
EDN3
αMSH
EDNRB
MC1R
Gα
Gαq
LRP5/6
ASIP
ATRN
cAMP
GRB2 SOS
NRAS-GDP
PKC
PI3K NF1 p120GAP PDK1
NRAS-GTP PKA
PKB/AKT
DSH
BRAF
GSK3β
MEK
CTTNB1
ERK
RAP
MITF
Figure 1. Clinical variablity in large CMN may be due to the many potential modifiers of a mosaic NRAS or BRAF gain-of-function genotype. (a) A 5-year-old girl with a typical giant CMN of the middle and lower back, flanks, abdomen, genitogluteal region, and upper thighs. After two years’ follow-up, she has no other syndromic features. Image reprinted with written permission from the family. (b) BRAF, mutated at V600 in exclusively small-to-medium-sized CMN, and NRAS, mutated at Q61 preferentially in large or giant-sized CMN (see Charbel et al. 2014), are at the nexus of multiple signaling networks depicted in this cartoon, that are important for melanoblast proliferation and melanocyte differentiation. This nexus coordinates the convergence of signals from receptor tyrosine kinase and G-protein-coupled receptors on not only MAPK-ERK but also PI3K-AKT and other target pathways, to lead to transcriptional modifications.
880
Journal of Investigative Dermatology (2014), Volume 134
COMMENTARY
mutations are known to cause the encoded enzyme to become constitutively active, and they have been found in many non-cutaneous cancers, perhaps driving cells to proliferate in certain environments. Corroborating initial observations in Belgian (Dessars et al., 2009) and Chinese populations (Wu et al., 2011), Charbel et al. (2014) broke down known medium and large CMN by size in their large French cohort to demonstrate NRAS mutations exclusively in the largest malformations, and some BRAF mutations in the medium-sized malformations (o20 cm PAS) and in one small CMN. The NRAS mutations were always Q61K or Q61R; other melanoma-associated mutations such as Q61H, Q61L, G12D, and G12V have not been reported in CMN, whereas G13R has been reported only once (Dessars et al., 2009). These results are compatible with earlier findings in smaller CMN and acquired nevi, with an overall preponderance of the same NRAS mutations in small-to-medium CMN. BRAF-mutated medium CMN can be as large or larger than NRASmutated lesions, meaning that size alone does not predict genotype, except that the large and giant CMN reported by Charbel et al. (2014) and Kinsler et al. (2013) never seem to have BRAF mutations. Other, not yet identified, gene mutations may also contribute less commonly to cause certain CMN. Phadke et al. (2011) reported a GNAQ Q209 mutation in a CMN, with no further details except that it also had a proliferative nodule, using a technology that requires 5–10% mutant allele frequency. Q209 and other hotspot mutations of GNAQ, encoding the GTP-binding protein G(q) subunit alpha (Figure 1b), are nearly exclusively found in uveal melanoma and blue nevi. However, another possibility is that causative mosaic NRAS mutations sometimes escape detection. In support of this hypothesis, Charbel et al. (2014) had recourse first to Sanger sequencing (detection efficiency as low as 15% mutant allele load), then to targeted pyrosequencing and high-resolution melting analysis for NRAS exon 3 (both at approximately 5% detection
threshold for a mutant allele). Finally, working with two negative samples, they used a PCR-based system called enhanced (E) ice-COLD-PCR. This system exploits a chemically modified, complementary oligonucleotide to a given wild-type allele to enhance preferentially the amplification of mutated alleles at a theoretical frequency as low as 0.1%, although Charbel et al. (2014) were able to quantify by exome sequencing in one of the two CMN that the mutant NRAS allele load was on the order of 1%. Large/giant CMN are usually sporadic, but small CMN are found more frequently in close relatives than in the general population. Siblings and first cousins of patients with large CMN have also been reported on rare occasion to have large CMN. A predisposing background of autosomal dominant inheritance of other genetic material may explain these unusual familial recurrences. However, given the inability of multiple exome sequencing studies to reproducibly identify any but known somatic mutations in the affected CMN tissue itself (Charbel et al., 2014, and personal communications), such predispositions are most likely the result of non-coding mutations, and therefore, they remain difficult to identify. Further correlations between a specific amino acid substitution and outcome or syndromic features can only be drawn if investigators continue to identify the entire spectrum of somatic mutations in CMN by using recently described highly sensitive detection techniques (Charbel et al., 2014; Kinsler et al., 2013), and to correlate them with detailed phenotypic information. Etiological implications
Both CMN and normal melanocytes arise from an embryonic neural crest cell population that separates from the future CNS before the end of the first month of gestation. As these cells multiply, they colonize all the tissues of the body. In mature human skin, descendant melanoblasts reside in the epidermis and sparsely populate the dermis as well as many non-cutaneous sites, such as the valves of the heart and the meninges. Constitutive activation of
NRAS or BRAF within a human neural crest cell may drive the prenatal proliferation of embryonic melanoblasts, but leads to a generally senescent CMN by birth. The best published models to test this hypothesis have been genetically engineered mice in which constitutively activated Nras is restricted to certain CNS and all pigment cell lineages from midgestation on. Soon after birth, the skin develops generalized, nevus-like dermal melanocytosis. Spontaneous melanomas form in a proportion of animals during their adulthood, leading to their premature death. Depending on the amino acid mutated, the melanomas can be cutaneous but include rarer acral and genital sites, or arise directly and only in the CNS out of presymptomatic, meningeal NCM. In contrast, similar conditional activation of even a single allele of Braf V600E in mice leads to prenatal death in association with dramatic brain, heart, and eye malformations (Dhomen et al., 2009). Despite featuring mosaic expression of MAPK pathway genes with the same mutations as found in both human CMN and melanoma, none of these mice develop lesions that greatly resemble human giant or multiple medium CMN. This may be in part the result of the genetic tools chosen to drive expression in all pigment cells, in order to investigate melanomagenesis. More adapted models for CMN are currently under investigation. A complementary hypothesis for nevus etiology is that smaller lesions arise from the descendants of a single somatically mutated cell, where its driver mutation had happened later than in the larger lesions. On the basis of the mouse models, if heterozygous, activating mutations of BRAF had occurred in the melanoblast lineage or its earlier progenitors, such mutations would have been incompatible with postnatal life. In contrast, NRAS mutations acquired in multipotent lineages, such as a neural crest cell or an earlier neuroepithelial cell, also giving rise to neural crest, are compatible with survival to birth and beyond. Such patients would have more extensive CMN and potentially carry NRAS mutations in additional tissues, whereas only smaller (later) CMN can harbor BRAF mutations. www.jidonline.org
881
COMMENTARY
Why the increased incidence of melanoma in young CMN patients? An example of a postnatally acquired Q61K mutation in the second copy of the NRAS gene (or loss of the normal allele) in a CMN cell has been correlated to the onset of malignant melanoma in one patient; similar loss of heterozygosity of NRAS Q61R was linked to nevomelanocyte proliferation within the CNS, causing NCM in another (Kinsler et al., 2013). However, heterozygous Q61R and Q61K NRAS mutations have been observed in primary CNS melanomas developing in two children with CMN, where the same heterozygous NRAS mutation had also been identified in the CMN itself (Pedersen et al., 2013). Curiously, no reliable case of melanoma arising within an associated small CMN has yet been reported, although these ‘‘satellite’’ lesions bear the same NRAS mutations as the primary, large CMN, and they are often exposed to sunlight (Charbel et al., 2014; Kinsler et al., 2013). Also, intriguingly, there are no reports of any coincident BRAF and NRAS mutations in the same CMN, despite the highly sensitive detection techniques that have been used. Unlike blue nevi, the most superficial component of a CMN is the most highly pigmented, conferring brown-to-black shades to the overlying epidermis. During childhood, while the skin continues to mature postnatally, CMN can evolve in hue or surface texture. CMN may also have congenitally fewer sweat glands than unaffected skin, resulting in potential overheating episodes or increased sweating in other areas of the body to compensate. Areas of larger CMN may have notably less fat under the skin, particularly around the flanks, limbs and buttocks. The CMN-involved epidermis may be dry or prone to atopic dermatitis in correlation with disorganized or absent sebaceous glands. Intermittent or chronic severe pruritis and mast cell infiltration of large CMN have also been reported at a recent conference (http://cmnexperts.org). It remains unclear whether the somatic mutations identified in human CMN are restricted to melanoblasts or may also extend to other neural crest or even neuroepithelial derivatives, but the latter scenario appears likely. Interestingly, targeting of Braf V600E to specific sensory 882
neurons of the mouse dorsal root ganglia, whose axons harbor unpigmented melanoblasts and which are also derived from neural crest cells, changes their fate, leading to overexpression of protein mediators of chronic itch and to intractable scratching behavior (Zhao et al., 2013). If the somatic mutations found in CMN are also present in or around the cells of peripheral nerves, using sensitive detection technologies such as those employed by Charbel et al. (2014) and others will have important and perhaps unexpected implications for management. CONFLICT OF INTEREST
The author states no conflict of interest.
REFERENCES Charbel C, Fontaine RH, Malouf GG et al. (2014) NRAS mutation is the sole recurrent somatic mutation in large congenital melanocytic nevi. J Invest Dermatol 134: 1067–74 Dessars B, De Raeve LE, Morandini R et al. (2009) Genotypic and gene expression studies in congenital melanocytic nevi: insight into initial steps of melanotumorigenesis. J Invest Dermatol 129:139–47 Dhomen N, Da Rocha Dias S, Hayward R et al. (2009) Inducible expression of (V600E) Braf
using tyrosinase-driven Cre recombinase results in embryonic lethality. Pigment Cell Melanoma Res 23:112–20 Kinsler VA, Thomas AC, Ishida M et al. (2013) Multiple congenital melanocytic nevi and neurocutaneous melanosis are caused by postzygotic mutations in codon 61 of NRAS. J Invest Dermatol 133:2229–36 Krengel S, Scope A, Dusza SW et al. (2013) New recommendations for the categorization of cutaneous features of congenital melanocytic nevi. J Am Acad Dermatol 68:441–51 Pedersen M, Ku¨sters-Vandevelde HVN, Viros A et al. (2013) Primary melanoma of the CNS in children is driven by congenital expression of oncogenic NRAS in melanocytes. Cancer Discov 3:458–69 Phadke PA, Rakheja D, Le LP et al. (2011) Proliferative nodules arising within congenital melanocytic nevi: a histologic, immunohistochemical, and molecular analyses of 43 cases. Am J Surg Pathol 35:656–69 Shah KN (2010) The risk of melanoma and neurocutaneous melanosis associated with congenital melanocytic nevi. Semin Cutan Med Surg 29:159–64 Wu D, Wang M, Wang X et al. (2011) Lack of BRAF(V600E) mutations in giant congenital melanocytic nevi in a Chinese population. Am J Dermatopathol 33:341–4 Zhao Z, Huo F, Jeffry J et al. (2013) Chronic itch development in sensory neurons requires BRAF signaling pathways. J Clin Invest 123:4769–80
See related article on pg 1119
A Fibronectin-Derived Cell Survival Peptide Belongs to a New Class of Epiviosamines Tomoo Ohashi1 In this issue, Lin et al. report the discovery of P12, a 14 amino acid peptide from the first fibronectin (FN) type III domain of FN, which has the capability of enhancing cell survival in culture and improving wound healing in rat skin. P12 belongs to a new class of bioactive peptides that they have named epiviosamines. Epiviosamines may have clinical applications. Journal of Investigative Dermatology (2014) 134, 882–884; doi:10.1038/jid.2013.470
Fibronectin (FN) is an extracellular matrix (ECM) protein. It can bind a variety of integrins on the cell surface
as well as many other molecules such as gelatin (denatured collagen), fibrin (fibrinogen), heparin/heparan sulfate, tissue
1
Department of Cell Biology, Duke University Medical Center, Durham, North Carolina, USA
Correspondence: Tomoo Ohashi, Department of Cell Biology, Duke University Medical Center, Box 3709, Durham, North Carolina 27710, USA. E-mail: [email protected]
Journal of Investigative Dermatology (2014), Volume 134
Tarafder AK, Bolasco G, Correia MS et al. (2014) Rab11b mediates melanin transfer between donor melanocytes and acceptor keratinocytes via coupled exo/endocytosis. J Invest Dermatol 135:1056–66 Van Den Bossche K, Naeyaert JM et al. (2006) The quest for the mechanism of melanin transfer. Traffic 7:769–78 Wasmeier C, Romao M, Plowright L et al. (2006) Rab38 and Rab32 control post-Golgi trafficking of melanogenic enzymes. J Cell Biol 175:271–81
Weiner L, Han R, Scicchitano BM et al. (2007) Dedicated epithelial recipient cells determine pigmentation patterns. Cell 130: 932–42 Wu XS, Masedunskas A, Weigert R et al. (2012) Melanoregulin regulates a shedding mechanism that drives melanosome transfer from melanocytes to keratinocytes. Proc Natl Acad Sci USA 109:E2101–9 Yamaguchi Y, Hearing VJ (2009) Physiological factors that regulate skin pigmentation. Biofactors 35:193–9
See related article on pg 1067
Hiding in Plain Sight: Molecular Genetics Applied to Giant Congenital Melanocytic Nevi Heather C. Etchevers1 Large and giant congenital melanocytic nevi are rare malformations that offer surprising insight into prenatal and postnatal acquisition of nevi of any size, central and peripheral nervous system development, and melanomagenesis. In this issue, Charbel et al. demonstrate the use of highly sensitive detection techniques for recurrent but difficult-to-detect mutations in NRAS and BRAF. It is now possible to systematically add a molecular qualifier to distinguish lesions that had once been considered to be equivalent based on the single visual parameter of size. These findings help to elucidate the pathophysiology of congenital melanocytic nevi and their predisposition to malignancy. Journal of Investigative Dermatology (2014) 134, 879–882; doi:10.1038/jid.2013.531
Nosology of large and giant congenital melanocytic nevi
Congenital melanocytic nevi (CMN) are malformations resulting from the faulty development of melanocyte progenitors in the embryo or fetus, and they are composed ultimately of an abnormal mixture of skin cellular elements. These pigmented hamartomas occupy sharply defined areas along the epidermal– dermal junction that range from a few millimeters in diameter to large swathes of the body, limbs or head. In the larger forms, CMN (single or multiple) often extend vertically into the deeper dermis and more rarely into the hypodermis or even subcutaneous tissues. The first descriptions of children with large CMN date from observations recorded by the French Count of Buffon in 1777, but the
incidence of CMN seems to be independent of ethnic factors. CMN have been classified historically according to their predicted largest diameter in adulthood, as if they were circular (predicted adult size (PAS), as they enlarge proportionately to the child’s growth). Until recently, other qualifiers had not systematically been taken into account. CMN measuring 40 cm or larger in PAS are referred to as ‘‘giant’’ nevi, according to the most recent and complete classification system (Krengel et al., 2013). This is intended to promote an international standard in phenotyping by including additional criteria such as pilosity, color heterogeneity, rugosity, and the presence of nodular growths, with charts for predicting PAS according
1
INSERM, UMR_S910, Aix-Marseille Universite´, Marseille, France
Correspondence: Heather C. Etchevers, INSERM, UMR_S910, Aix-Marseille Universite´, Faculte´ de Me´decine, 27 Boulevard Jean Moulin, Marseille 13005, France. E-mail: [email protected]
to body site and photographic examples. Small (o1.5 cm PAS) CMN occur in more than 1 in 100 births. Larger CMN (Figure 1a) form a much rarer subset, with prevalence estimated at around 1 in 20,000 to 50,000 births as a function of size. Treatment options in the year 2014 remein exclusively surgical. The differential diagnosis of small and medium CMN includes smooth muscle hamartoma or Becker’s nevus, mastocytoma, variants of dermal melanocytosis, and cafe´-au-lait macules. Larger CMN have been confused with pigmented plexiform neurofibromas. Histologic evaluation, dermoscopic evaluation, and the development of typical CMN features over time may clarify the diagnosis. However, molecular characterization promises to be an important additional criterion in phenotyping even the most easily diagnosed lesions and in teasing prognostic factors for syndromic features such as symptomatic neurocutaneous melanocytosis or malignant melanoma. Sensitive molecular diagnoses, such as those most recently developed in this issue of JID by Charbel et al. (2014), may soon augment the diagnostic and therapeutic arsenal. ‘‘Satellite’’ is a term used commonly to describe small or medium CMN, or those that appear postnatally (‘‘tardive’’ nevi) in the presence of a large/giant CMN. Both semantically and molecularly, it is more accurate to refer to ‘‘disseminated’’ nevi (Kinsler et al., 2013). These disseminated CMN may be present at birth and/or may increase to significant numbers over the first few years of life. Occasionally, a single largest CMN cannot be considered to be the principal malformation among the many, in which case the child is said to have ‘‘multiple medium’’ CMN. This presentation correlates with an increased predisposition to neurological abnormalities, as are greater satellite numbers and PAS420 cm (large and giant CMN; Shah, 2010). Contrary to a common interpretation of the earliest reports of such correlations, there is no discrete biological cut-off as to how many additional disseminated CMN predispose to neurological symptoms or melanoma. www.jidonline.org
879
COMMENTARY
Clinical Implications
Gain-of-function somatic mutations in NRAS at Q61 have been found in large and giant congenital melanocytic nevi (CMN), but mutant alleles can be difficult to detect.
Gain-of-function somatic mutations in either BRAF at V600 or NRAS Q61 have been found in small and medium CMN.
The principal CMN harbors the same mutation as found in samples from disseminated nevi or complications such as melanoma or central nervous system anomalies of the patient, implying their common origin.
within the malformation, but it is often extracutaneous, with some predilection for the CNS. Other malignancies such as liposarcomas, rhabdomyosarcomas, and peripheral nerve sheath tumors have also been described in association with large CMN. Rapidly growing ‘proliferative nodules’ within the CMN can mimic melanoma but often show benign features on closer examination (Phadke et al., 2011). Molecular basis of nevus formation
Syndromic CMN
Neurocutaneous melanocytosis, cited in earlier literature as neurocutaneous melanosis (NCM), is characterized by abnormal aggregations of nevomelanocytes within the central nervous system (CNS) of an estimated 5–15% of patients with the larger forms of CMN, or multiple medium-sized CMN. Neurological signs may include hydrocephalus, epilepsy, arachnoid cysts, tethered spinal cord, Dandy–Walker malformation (cerebellar vermis hypoplasia), developmental delay, and a number of rarer CNS tumors (Kinsler et al., 2013). True melanosis, through accumulation of
Both CMN and acquired melanocytic nevi are associated exclusively with somatic mutations in intracellular proteins of the microtubule-associated protein kinase (MAPK) signal transduction pathway (Figure 1b). In some series of CMN, only a recurrent mutation in the NRAS gene at codon 61 has been described, transforming the glutamine (Q) to lysine (K) or arginine (R) (Kinsler et al., 2013, and references therein), but many others have described a single hotspot mutation in BRAF (V600E) as well (Dessars et al., 2009; Phadke et al., 2011)—as is found commonly in adultonset melanoma. All of the identified
melanin in CNS neurons or glia, may also occur in association with CMN, but further neuropathological studies need to be undertaken for this to be conclusive. Melanoma develops in an estimated 1–2% of pre-pubertal patients with large CMN or NCM, on the order of 10,000 times more often than the incidence among the general population of children under the age of 10, based on registry data from the SEER (Surveillance, Epidemiology, and End Results) program of the US National Cancer Institute, 1975–2000. CMN-associated melanoma may be cutaneous, usually
a
b DKK1 WNT
FZ
KITL
FGF
EGF
KIT
FGFR
EGFR
HGF
MET
EDN3
αMSH
EDNRB
MC1R
Gα
Gαq
LRP5/6
ASIP
ATRN
cAMP
GRB2 SOS
NRAS-GDP
PKC
PI3K NF1 p120GAP PDK1
NRAS-GTP PKA
PKB/AKT
DSH
BRAF
GSK3β
MEK
CTTNB1
ERK
RAP
MITF
Figure 1. Clinical variablity in large CMN may be due to the many potential modifiers of a mosaic NRAS or BRAF gain-of-function genotype. (a) A 5-year-old girl with a typical giant CMN of the middle and lower back, flanks, abdomen, genitogluteal region, and upper thighs. After two years’ follow-up, she has no other syndromic features. Image reprinted with written permission from the family. (b) BRAF, mutated at V600 in exclusively small-to-medium-sized CMN, and NRAS, mutated at Q61 preferentially in large or giant-sized CMN (see Charbel et al. 2014), are at the nexus of multiple signaling networks depicted in this cartoon, that are important for melanoblast proliferation and melanocyte differentiation. This nexus coordinates the convergence of signals from receptor tyrosine kinase and G-protein-coupled receptors on not only MAPK-ERK but also PI3K-AKT and other target pathways, to lead to transcriptional modifications.
880
Journal of Investigative Dermatology (2014), Volume 134
COMMENTARY
mutations are known to cause the encoded enzyme to become constitutively active, and they have been found in many non-cutaneous cancers, perhaps driving cells to proliferate in certain environments. Corroborating initial observations in Belgian (Dessars et al., 2009) and Chinese populations (Wu et al., 2011), Charbel et al. (2014) broke down known medium and large CMN by size in their large French cohort to demonstrate NRAS mutations exclusively in the largest malformations, and some BRAF mutations in the medium-sized malformations (o20 cm PAS) and in one small CMN. The NRAS mutations were always Q61K or Q61R; other melanoma-associated mutations such as Q61H, Q61L, G12D, and G12V have not been reported in CMN, whereas G13R has been reported only once (Dessars et al., 2009). These results are compatible with earlier findings in smaller CMN and acquired nevi, with an overall preponderance of the same NRAS mutations in small-to-medium CMN. BRAF-mutated medium CMN can be as large or larger than NRASmutated lesions, meaning that size alone does not predict genotype, except that the large and giant CMN reported by Charbel et al. (2014) and Kinsler et al. (2013) never seem to have BRAF mutations. Other, not yet identified, gene mutations may also contribute less commonly to cause certain CMN. Phadke et al. (2011) reported a GNAQ Q209 mutation in a CMN, with no further details except that it also had a proliferative nodule, using a technology that requires 5–10% mutant allele frequency. Q209 and other hotspot mutations of GNAQ, encoding the GTP-binding protein G(q) subunit alpha (Figure 1b), are nearly exclusively found in uveal melanoma and blue nevi. However, another possibility is that causative mosaic NRAS mutations sometimes escape detection. In support of this hypothesis, Charbel et al. (2014) had recourse first to Sanger sequencing (detection efficiency as low as 15% mutant allele load), then to targeted pyrosequencing and high-resolution melting analysis for NRAS exon 3 (both at approximately 5% detection
threshold for a mutant allele). Finally, working with two negative samples, they used a PCR-based system called enhanced (E) ice-COLD-PCR. This system exploits a chemically modified, complementary oligonucleotide to a given wild-type allele to enhance preferentially the amplification of mutated alleles at a theoretical frequency as low as 0.1%, although Charbel et al. (2014) were able to quantify by exome sequencing in one of the two CMN that the mutant NRAS allele load was on the order of 1%. Large/giant CMN are usually sporadic, but small CMN are found more frequently in close relatives than in the general population. Siblings and first cousins of patients with large CMN have also been reported on rare occasion to have large CMN. A predisposing background of autosomal dominant inheritance of other genetic material may explain these unusual familial recurrences. However, given the inability of multiple exome sequencing studies to reproducibly identify any but known somatic mutations in the affected CMN tissue itself (Charbel et al., 2014, and personal communications), such predispositions are most likely the result of non-coding mutations, and therefore, they remain difficult to identify. Further correlations between a specific amino acid substitution and outcome or syndromic features can only be drawn if investigators continue to identify the entire spectrum of somatic mutations in CMN by using recently described highly sensitive detection techniques (Charbel et al., 2014; Kinsler et al., 2013), and to correlate them with detailed phenotypic information. Etiological implications
Both CMN and normal melanocytes arise from an embryonic neural crest cell population that separates from the future CNS before the end of the first month of gestation. As these cells multiply, they colonize all the tissues of the body. In mature human skin, descendant melanoblasts reside in the epidermis and sparsely populate the dermis as well as many non-cutaneous sites, such as the valves of the heart and the meninges. Constitutive activation of
NRAS or BRAF within a human neural crest cell may drive the prenatal proliferation of embryonic melanoblasts, but leads to a generally senescent CMN by birth. The best published models to test this hypothesis have been genetically engineered mice in which constitutively activated Nras is restricted to certain CNS and all pigment cell lineages from midgestation on. Soon after birth, the skin develops generalized, nevus-like dermal melanocytosis. Spontaneous melanomas form in a proportion of animals during their adulthood, leading to their premature death. Depending on the amino acid mutated, the melanomas can be cutaneous but include rarer acral and genital sites, or arise directly and only in the CNS out of presymptomatic, meningeal NCM. In contrast, similar conditional activation of even a single allele of Braf V600E in mice leads to prenatal death in association with dramatic brain, heart, and eye malformations (Dhomen et al., 2009). Despite featuring mosaic expression of MAPK pathway genes with the same mutations as found in both human CMN and melanoma, none of these mice develop lesions that greatly resemble human giant or multiple medium CMN. This may be in part the result of the genetic tools chosen to drive expression in all pigment cells, in order to investigate melanomagenesis. More adapted models for CMN are currently under investigation. A complementary hypothesis for nevus etiology is that smaller lesions arise from the descendants of a single somatically mutated cell, where its driver mutation had happened later than in the larger lesions. On the basis of the mouse models, if heterozygous, activating mutations of BRAF had occurred in the melanoblast lineage or its earlier progenitors, such mutations would have been incompatible with postnatal life. In contrast, NRAS mutations acquired in multipotent lineages, such as a neural crest cell or an earlier neuroepithelial cell, also giving rise to neural crest, are compatible with survival to birth and beyond. Such patients would have more extensive CMN and potentially carry NRAS mutations in additional tissues, whereas only smaller (later) CMN can harbor BRAF mutations. www.jidonline.org
881
COMMENTARY
Why the increased incidence of melanoma in young CMN patients? An example of a postnatally acquired Q61K mutation in the second copy of the NRAS gene (or loss of the normal allele) in a CMN cell has been correlated to the onset of malignant melanoma in one patient; similar loss of heterozygosity of NRAS Q61R was linked to nevomelanocyte proliferation within the CNS, causing NCM in another (Kinsler et al., 2013). However, heterozygous Q61R and Q61K NRAS mutations have been observed in primary CNS melanomas developing in two children with CMN, where the same heterozygous NRAS mutation had also been identified in the CMN itself (Pedersen et al., 2013). Curiously, no reliable case of melanoma arising within an associated small CMN has yet been reported, although these ‘‘satellite’’ lesions bear the same NRAS mutations as the primary, large CMN, and they are often exposed to sunlight (Charbel et al., 2014; Kinsler et al., 2013). Also, intriguingly, there are no reports of any coincident BRAF and NRAS mutations in the same CMN, despite the highly sensitive detection techniques that have been used. Unlike blue nevi, the most superficial component of a CMN is the most highly pigmented, conferring brown-to-black shades to the overlying epidermis. During childhood, while the skin continues to mature postnatally, CMN can evolve in hue or surface texture. CMN may also have congenitally fewer sweat glands than unaffected skin, resulting in potential overheating episodes or increased sweating in other areas of the body to compensate. Areas of larger CMN may have notably less fat under the skin, particularly around the flanks, limbs and buttocks. The CMN-involved epidermis may be dry or prone to atopic dermatitis in correlation with disorganized or absent sebaceous glands. Intermittent or chronic severe pruritis and mast cell infiltration of large CMN have also been reported at a recent conference (http://cmnexperts.org). It remains unclear whether the somatic mutations identified in human CMN are restricted to melanoblasts or may also extend to other neural crest or even neuroepithelial derivatives, but the latter scenario appears likely. Interestingly, targeting of Braf V600E to specific sensory 882
neurons of the mouse dorsal root ganglia, whose axons harbor unpigmented melanoblasts and which are also derived from neural crest cells, changes their fate, leading to overexpression of protein mediators of chronic itch and to intractable scratching behavior (Zhao et al., 2013). If the somatic mutations found in CMN are also present in or around the cells of peripheral nerves, using sensitive detection technologies such as those employed by Charbel et al. (2014) and others will have important and perhaps unexpected implications for management. CONFLICT OF INTEREST
The author states no conflict of interest.
REFERENCES Charbel C, Fontaine RH, Malouf GG et al. (2014) NRAS mutation is the sole recurrent somatic mutation in large congenital melanocytic nevi. J Invest Dermatol 134: 1067–74 Dessars B, De Raeve LE, Morandini R et al. (2009) Genotypic and gene expression studies in congenital melanocytic nevi: insight into initial steps of melanotumorigenesis. J Invest Dermatol 129:139–47 Dhomen N, Da Rocha Dias S, Hayward R et al. (2009) Inducible expression of (V600E) Braf
using tyrosinase-driven Cre recombinase results in embryonic lethality. Pigment Cell Melanoma Res 23:112–20 Kinsler VA, Thomas AC, Ishida M et al. (2013) Multiple congenital melanocytic nevi and neurocutaneous melanosis are caused by postzygotic mutations in codon 61 of NRAS. J Invest Dermatol 133:2229–36 Krengel S, Scope A, Dusza SW et al. (2013) New recommendations for the categorization of cutaneous features of congenital melanocytic nevi. J Am Acad Dermatol 68:441–51 Pedersen M, Ku¨sters-Vandevelde HVN, Viros A et al. (2013) Primary melanoma of the CNS in children is driven by congenital expression of oncogenic NRAS in melanocytes. Cancer Discov 3:458–69 Phadke PA, Rakheja D, Le LP et al. (2011) Proliferative nodules arising within congenital melanocytic nevi: a histologic, immunohistochemical, and molecular analyses of 43 cases. Am J Surg Pathol 35:656–69 Shah KN (2010) The risk of melanoma and neurocutaneous melanosis associated with congenital melanocytic nevi. Semin Cutan Med Surg 29:159–64 Wu D, Wang M, Wang X et al. (2011) Lack of BRAF(V600E) mutations in giant congenital melanocytic nevi in a Chinese population. Am J Dermatopathol 33:341–4 Zhao Z, Huo F, Jeffry J et al. (2013) Chronic itch development in sensory neurons requires BRAF signaling pathways. J Clin Invest 123:4769–80
See related article on pg 1119
A Fibronectin-Derived Cell Survival Peptide Belongs to a New Class of Epiviosamines Tomoo Ohashi1 In this issue, Lin et al. report the discovery of P12, a 14 amino acid peptide from the first fibronectin (FN) type III domain of FN, which has the capability of enhancing cell survival in culture and improving wound healing in rat skin. P12 belongs to a new class of bioactive peptides that they have named epiviosamines. Epiviosamines may have clinical applications. Journal of Investigative Dermatology (2014) 134, 882–884; doi:10.1038/jid.2013.470
Fibronectin (FN) is an extracellular matrix (ECM) protein. It can bind a variety of integrins on the cell surface
as well as many other molecules such as gelatin (denatured collagen), fibrin (fibrinogen), heparin/heparan sulfate, tissue
1
Department of Cell Biology, Duke University Medical Center, Durham, North Carolina, USA
Correspondence: Tomoo Ohashi, Department of Cell Biology, Duke University Medical Center, Box 3709, Durham, North Carolina 27710, USA. E-mail: [email protected]
Journal of Investigative Dermatology (2014), Volume 134
