Phenotypic Characterization of Nevus and Tumor Patterns in MITF E318K Mutation Carrier Melanoma Patients

Transcription

1 See related commentary on pg 16 ORIGINAL ARTICLE Phenotypic Characterization of Nevus and Tumor Patterns in MITF E318K Mutation Carrier Melanoma Patients Richard A. Sturm 1,CarlyFox 2, Phil McClenahan 3, Kasturee Jagirdar 1, Maider Ibarrola-Villava 1,4, Parastoo Banan 3,5,NicolaC.Abbott 3,5, Gloria Ribas 4, Brian Gabrielli 2, David L. Duffy 6 and H. Peter Soyer 3,5 A germline polymorphism of the microphthalmia transcription factor (MITF) gene encoding a SUMOylationdeficient E318K-mutated protein has recently been described as a medium-penetrance melanoma gene. In a clinical assessment of nevi from 301 volunteers taken from Queensland, we identified six individuals as MITF E318K mutation carriers. The phenotype for 5 of these individuals showed a commonality of fair skin, body freckling that varied over a wide range, and total nevus count between 46 and 430; in addition, all were multiple primary melanoma patients. The predominant dermoscopic signature pattern of nevi was reticular, and the frequency of globular nevi in carriers varied, which does not suggest that the MITF E318K mutation acts to force the continuous growth of nevi. Excised melanocytic lesions were available for four MITF E318K carrier patients and were compared with a matched range of wild-type (WT) melanocytic lesions. The MITF staining pattern showed a predominant nuclear signal in all sections, with no significant difference in the nuclear/cytoplasmic ratio between mutation-positive or -negative samples. A high incidence of amelanotic melanomas was found within the group, with three of the five melanomas from one patient suggesting a genetic interaction between the MITF E318K allele and an MC1R homozygous red hair color (RHC) variant genotype. Journal of Investigative Dermatology (2014) 134, ; doi: /jid ; published online 18 July 2013 INTRODUCTION The microphthalmia transcription factor (MITF) gene plays a crucial role in melanocyte growth, development, and pigmentation (Cheli et al., 2010). In addition, it has been proposed to act as a lineage survival oncogene because of gene amplification in metastatic melanoma (Garraway et al., 2005). Several different somatic mutations have been found in MITF in melanoma tumor samples (Cronin et al., 2009), and one recurrent germline mutation MITF E318K that predisposes to 1 Melanogenix Group, Institute for Molecular Bioscience, The University of Queensland, Brisbane, Queensland, Australia; 2 The University of Queensland Diamantina Institute, Translational Research Institute, Brisbane, Queensland, Australia; 3 Dermatology Research Centre, The University of Queensland, School of Medicine, Princess Alexandra Hospital, Brisbane, Queensland, Australia; 4 Department of Haematology and Medical Oncology, Fundacion Investigacion Hospital Clinico-INCLIVA, Valencia, Spain; 5 Department of Dermatology, Princess Alexandra Hospital, Brisbane, Queensland, Australia and 6 Queensland Institute of Medical Research, Brisbane, Queensland, Australia Correspondence: Richard A. Sturm, Institute for Molecular Bioscience, University of Queensland, 306 Carmody Road, St Lucia, Brisbane, Queensland, Australia. r.sturm@imb.uq.edu.au Abbreviations: APE1, apurinic/apyrimidinic endonuclease 1; CDK2, cyclindependent kinase 2; CDK4, cyclin-dependent kinase 4; CDKN2A, cyclindependent kinase inhibitor 2A; IHC, immunohistochemistry; MC1R, melanocortin-1 receptor; MITF, microphthalmia transcription factor; RHC, red hair color; WT, wild type Received 3 February 2013; revised 23 April 2013; accepted 20 May 2013; accepted article preview online 17 June 2013; published online 18 July 2013 melanoma and renal cell carcinoma has recently been described in independent French and Australian studies (Bertolotto et al., 2011; Yokoyama et al., 2011). This variant allele encodes a protein that biochemically impairs MITF SUMOylation at position 316K and is genetically associated with increased nevus number, non-blue eye color, over a fivefold increased risk for melanoma, and higher incidence of multiple primary melanomas. The MITF E318K mutation has also been studied in a group of Italian melanoma patients (Ghiorzo et al., 2013) in whom a higher risk for melanoma and the development of multiple primary melanomas were again reported; notably, in this group a predisposition to pancreatic cancer was also found. On examination of the subtypes of melanoma in these cases, a high prevalence of nodular melanoma was described (Ghiorzo et al., 2013), and it has been suggested that the mutated MITF E318K protein may direct an altered transcription program in melanocytic cells (Bertolotto et al., 2011; Yokoyama et al., 2011) and manifest in characteristic histopathological or immunohistochemical (IHC) (Bertolotto et al., 2011) consequences for melanomas in carrier individuals. We have initiated a study of nevi in volunteers from the southeast Queensland area (Douglas et al., 2011; McClenahan et al., 2012) to determine the pigmentation phenotype, the nevus phenotype, and the dermoscopic nevus subtypes to combine this data set with genotypic comparisons of these individuals. In doing so, six individuals, all of whom were & 2014 The Society for Investigative Dermatology 141

2 multiple primary melanoma patients, were identified as MITF E318K carriers. We have used this sample set to examine whether there is any commonality in the pigmentation or nevus phenotype, including number, distribution, color, and size of nevi, as well as in the dermoscopic or IHC staining pattern of nevi and melanoma samples, that may distinguish these patients from the normal population and from other melanoma patients. RESULTS Identification and phenotypic characteristics of MITF mutation carrier melanoma patients compared with noncarrier melanoma patients We clinically assessed 288 of the 301 subjects in our study for a range of pigmentation characteristics and nevus phenotypes. This group consisted of 168 individuals with a personal history of melanoma, 31 with a first-degree relative with melanoma, 95 without such a history, and 7 unable to be classified. Several of the melanoma patient volunteers were not able to fully participate in the physical assessment protocol, and some of the controls who provided saliva samples were omitted; however, all participants were genotyped for a number of pigmentation and nevogenic genes (Duffy et al., 2010b). This included the rs *g/a MITF E318K polymorphism (Bertolotto et al., 2011; Yokoyama et al., 2011), with six individuals found to be heterozygous carriers of this allele in a cohort of five males and one female (Table 1). Each of these six individuals were screened and found to be without mutation at the CDKN2A locus. Two red-haired individuals, 149DC and 176SP, were melanocortin-1 receptor (MC1R) R/R homozygotes, as expected (Beaumont et al., 2011). There was only one IRF4 (IFN regulatory factor 4) rs *c/t carrier identified as 130MV who was a heterozygote; this single-nucleotide polymorphism has previously been associated with lower nevus counts in adults in the homozygous T/T state (Duffy et al., 2010a). Five of these six patients were phenotyped following the protocol described by Douglas et al. (2011). All these five individuals were fair skinned, and their body freckling scores varied over a wide range, being highest in patient 149DC who was red haired and an MC1R R/R homozygote carrier, an association that we have previously reported (Duffy et al., 2004). This patient also recorded the lowest number (46) of total nevi above 2 mm over the whole body in the group, with all others having much higher scores and reaching a maximum of 430 in 130MV (Table 1). Notably, all MITF E318K patients had suffered from multiple primary melanomas, ranging from 2 to 5 (Table 1), compared with a recurrence rate of 39.7% seen for the MITF WT melanoma patients (P ¼ 0.003); this figure is in line with that reported for other melanoma cohorts, ranging between 22 and 40% (Baxter et al., 2008; Salama et al., 2013). None of the control or unknown subjects were found to be carrying the MITF E318K polymorphism, with an allele frequency of in our 199 individuals at high risk for melanoma. Nevus and melanoma patterns of MITF mutation carrier patients The total body nevus counts of these patients were followed up by recording nevus distribution, size, profile, color, and dermoscopic overall morphology of nevi above 5 mm all over the body, and included those 42 mm on the back (Table 2). Although an increase in the mean of total nevi was apparent when comparing the MITF E318K carriers with the MITF wildtype (WT) melanoma patients, this became highly significant (P ¼ 0.008) when nevi 45 mm were considered with counts of 84.2 (95% confidence interval (CI) ) and 32.6 (95% CI ) for the two groups, respectively (Table 2). The major contribution to the difference in nevi 45 þ 2mm on the back were greater reticular nevus counts of 55.4 (95% CI ) and 20.3 (95% CI ; P ¼ 0.004), respectively. The predominant dermoscopic signature pattern of MITF E318K patients was reticular (130MV, 149DC, 155IN), and then nonspecific (158RP, 176SP). The site of melanoma, its type, and its depth ascertained from pathology reports indicated that the major site of occurrence in these patients was on the back (P ¼ 0.006), followed by the leg, arm, and abdomen, possibly indicating a propensity for UVR-exposed sites. Of the total number of 18 melanomas excised from this cohort, 8 were melanoma in situ, including one lentigo maligna, with the remaining 10 ranging in depth from 0.15 to 2.9 mm, with one outlier at 9 mm. The 20% incidence of melanoma arising from preexisting nevi in the MITF E318K carriers (Table 2) is similar to that reported in studies of other multiple melanoma patients (Murali et al., 2012). Further, there were four amelanotic melanomas identified (Table 1), three of them in 149DC, with body photographs and representative nevi of this patient shown in Figure 1, and one in 130MV (other patients are shown in Supplementary Figure S1 online). The incidence of amelanotic melanoma of 30.7% in MITF E318K carrier patients is much greater than the generally reported frequency of 2 8% (Jaimes et al., 2012; McClain et al., 2012). IHC analysis of nevi and melanoma tumors excised from MITF mutation carriers Excised melanocytic lesions and accompanying pathology analyses were available for four MITF E318K mutation carrier patients (130MV, 149DC, 176SP, and M004RR), and could be compared with a matched range of MITF WT melanocytic lesion samples. Any melanocytic lesion from the genotyped patients was considered for histopathological examination and IHC analysis, provided there was sufficient tissue to permit sectioning and IHC staining. The sample set of 39 lesions consisted of 19 from mutation carriers and 20 from WT patients (Table 3). Comparisons were made between benign and dysplastic nevi, as well as between four melanomas from mutation carriers (149DC and M004RR) and two from an MITF WT patient (17DC) who had also suffered from multiple primary melanomas. The staining patterns for MITF protein (Granter et al., 2002) in representative melanoma sections are shown in Figure 2, and a comparison is made between melanoma in situ (Figure 2a) and invasive melanoma (Figure 2b) for MITF E318K tumors with in situ melanoma (Figure 2c) for an MITF WT sample. An intense MITF nuclear staining pattern is apparent in all sections, with no significant difference seen in the degree of nuclear/cytoplasmic ratio between mutation-positive or -negative samples, consistent with transfection studies of other SUMOylation-deficient 142 Journal of Investigative Dermatology (2014), Volume 134

3 Table 1. Pigmentation phenotype and genotype characteristics of MITF E318K allele carriers versus MITF WT melanoma cases Sample code Age (range years) Sex Skin color 1 Hair color 3 (%) Reflectance 2 (%) Eye color 4 (%) Total freckling Total Total melanoma score 5 nevi 6 (recurrent%) AM (%) MC1R MC1R RHC IRF4 genotype 7 genotype 8 rs MV 41 M F 67.49/59.01 F/b Bl WT/I155T þ /r C/T 149DC 61 M F 63.06/63.04 R Bl D84E/R160W R/R C/C 155IN 53 M F 63.93/61.67 F/b Bl V60L/D84E r/r C/C 158RP 61 M F 61.03/59.05 LB Br WT/R151C þ /R C/C 176SP 57 F F 63.2/55.87 R Br R151C/R151C R/R C/C M004RR 66 M WT/WT þ / þ C/C MITF E318K Mean (95% CI) 56.5 (41 66) 83% M 17% F 100% F 0% M 0% O 63.7/59.7 ( / ) 40% F/b 40% R 20% LB 0% DB 0% B 60% Bl 0% G 40% Br 4.2 ( ) 238 ( ) 3.0 ( ) (100%) 0.8 ( ) (30.7%) 0.17 þ / þ 0.17 þ /R 0.17 þ /r 0.0 r/r 0.17 r/r 0.33 R/R 0.83 C/C 0.17 C/T 0.0 T/T MITF WT 11 Mean (95% CI) 52.5 (14 85) 52% M 48% F 77% F 21% M 2% O 63.0/55.8 ( / ) 22% F/b 16% R 35% LB 23% DB 5% B 60% B 28% G 12% B 4.2 ( ) 198 ( ) 1.9 ( ) (39.7%) P ¼ þ / þ þ /R þ /r r/r r/r R/R Abbreviations: AM, amelanotic melanoma; CI, confidence interval; IRF4, IFN regulatory factor 4; MITF, microphthalmia transcription factor; RHC, red hair color; WT, wild type. 1 Skin color: F, fair; M, medium; O, olive dark. 2 Reflectance measured on inner/outer upper arm. 3 Hair color: B, black; DB, dark brown; F/b, fair blonde; LB, light brown; R, red. 4 Eye color: Bl, blue; Br, brown; G, green/hazel. 5 Summation of none, 0; mild, 1; moderate, 2; and severe, 3, rated on the face, dorsum of right hand, and shoulders with total score between 0 and 9 (Duffy et al., 2004). 6 Total nevi number include all nevi 45 mm on all body sites and 42mm on the back, and 42 mm on the lower leg of females. 7 Genotypes derived by complete sequencing. 8 Genotype based on RHC alleles as þ, r, and R as defined in Beaumont et al. (2011), and given as frequency. 9 Not recorded. 10 Not available, patient has died of melanoma. 11 From 159 to 161 melanoma cases with phenotype C/C C/T T/T 143

4 Table 2. Nevus and melanoma phenotype characteristics of MITF E318K allele carriers versus MITF WT melanoma cases Sample code Nevi 45mm Nevi 45 þ 2mm on back Globular nevi 1 (%) Reticular nevi 1 (%) Homogenous/ nonspecific 1 (%) Signature pattern Melanoma (site, Clark level, specifics) Lesion depth (mm) Association with nevus 130MV (11.1) 137 (65.9) 46 (22.1) Reticular Left forearm, lentigo maligna (Oct 2009) NA N Back, Clark level IV amelanotic (Oct 2009) 2.9 N 149DC (38.2) 14 (41.2) 7 (20.6) Reticular Back, Clark level V (Jun 2003) 9 2 Back, Clark level II (Mar 2004) 0.4 Y arising in a junctional dysplastic nevus Back, Clark level III amelanotic (Jun 2004) 0.8 N Right shoulder, melanoma in situ amelanotic (Jan 2011) NA N Back, melanoma in situ amelanotic (Apr 2011) NA N 155IN (6.4) 79 (72.5) 22 (20.2) Reticular Back, melanoma in situ (Apr 2008) NA Back, melanoma in situ (Jan 2012) NA N 158RP (15.9) 21 (33.3) 30 (47.6) Nonspecific Left shoulder, Clark level II (Dec 2001) 0.2 Y preexisting nevus Forearm, melanoma in situ (Aug 2007) NA 176SP (11.9) 26 (20.6) 83 (65.9) Nonspecific Back, Clark level NA (Jun 2001) 0.15 Abdomen, Clark level NA (Jun 2001) 0.9 Left thigh, melanoma in situ (Feb 2004) NA Left leg, Clark level NA (Dec 2006) 0.6 Left knee, melanoma in situ (Feb 2007) NA M004RR 2 Right shin, Clark level NA (Oct 2002) 0.47 N Left shoulder, Clark level II (Oct 2010) 1.1 N MITF E318K Mean (95% CI) 84.2 ( ) 109 ( ) 13.6 ( ) (12.8) 55.4 ( ) (52.0) 37.6 ( ) (35.2) 60% Reticular 40% Nonspecific 0% Globular NA NA 20% MITF WT 3 Mean (95% CI) 32.6 ( ) 54.8 ( ) 7.0 ( ) (12.7) 20.3 ( ) (36.8) 27.8 ( ) (50.5) 69% Reticular 28% Nonspecific 3% Globular NA NA P-value Abbreviations: AM, amelanotic melanoma; MITF, microphthalmia transcription factor; NA, not applicable; WT, wild type. 1 Nevi able to be classified. 2 Not available. 3 From 159 to 161 melanoma cases with phenotype. 4 Log-transformed mole count after adjusting for age, results of the Kruskal Wallis test and the w 2 test. 144 Journal of Investigative Dermatology (2014), Volume 134

5 Figure 1. Total body and dermoscopic images from patient 149DC. The photograph of the back of patient 149DC is shown together with nine dermoscopic images chosen as representative of the nevi in this patient who has a reticular signature pattern. A total of 34 nevi 45 mm in diameter on all body sites and 42mm in diameter on the back were identified on this patient. Excluding small nevi on the back results in 28 nevi 45 mm on the whole body. Of the 15 nevi on the back (median size 5.1 mm), 6 are 42 mm but o5 mm in diameter and 9 are 45 mm in diameter. Of these, 12 are reticular and 3 homogenous/unspecific. In addition, 2 are pink in color, 11 are light brown, and 2 are mid-brown. On other body sites, 1 nevus was identified on the head/neck, 4 on the abdomen/chest, 4 on the upper limbs, and 10 on the lower limbs. Nevus color was either pink or light brown for all of them. Table 3. IHC staining patterns of nevi and melanomas excised from MITF E318K allele 1 carriers and MITF WT 2 patients Intensity 3,4 Nuclear Cytoplasmic E318K WT P-value 5 E318K WT P-value 5 E318K WT P-value 5 MITF 2.26 (0.55) 2.3 (0.66) /19 20/20 1 3/19 9/ APE (0.9) 2.65 (0.67) /19 20/ /19 16/ CDK (0.88) 2.4 (0.75) /19 12/ /19 20/ Ki Abbreviations: APE1 apurinic/apyrimidinic endonuclease 1; CDK2, cyclin-dependent kinase 2; IHC, immunohistochemistry; MITF, microphthalmia transcription factor; WT, wild type. 1 Sample set included 4 benign nevi, 11 dysplastic nevi, 2 melanomas in situ, and 2 primary melanomas. 2 Sample set included 3 benign nevi, 15 dysplastic nevi, 1 melanoma in situ, and 1 primary melanoma. 3 Intensity was scored on a three-point scale (1 3) per slide and represented as the mean (±SD) of these values. 4 As an average of the percentage of tumor cells stained within the lesions, range was 1 10%, with none expressing 0%. 5 Fisher s exact test. 6 Mann Whitney test. mutations of the MITF protein (Murakami and Arnheiter, 2005), nor between stages. However, nests were seen to express the highest density of immunoreactivity, and MITF E318K tumors had fewer samples with cytoplasmic staining (only 3 of 19). This similarity of staining patterns between mutant and WT melanocytic lesions was also seen using two MITF target proteins, apurinic/apyrimidinic endonuclease 1 (APE1) (Liu et al., 2009) and cyclin-dependent kinase 2 (CDK2; Table 3) (Du et al., 2004). Further, there was no difference in the average proliferation index as measured by Ki67 staining of comparative sections; however, the invasive melanomas (as shown in Figure 2b) demonstrated the greatest level of reactivity (10%). DISCUSSION There are now a multitude of genetic loci that are known to influence melanoma risk (Law et al., 2012), but it is imperative to understand those that influence the number and behavior of nevi as these confer the greatest individual risk of melanomagenesis. In this study the germline missense substitution MITF E318K, known as a medium-penetrance melanoma gene, was found only in our melanoma patient cohort of 168 individuals, excluding other at-risk individuals, giving an allele frequency of This is in line with other reports of this mutant allele in melanoma cases in Australian/UK (Yokoyama et al., 2011), in French (Bertolotto et al., 2011), and in Italian samples (Ghiorzo et al., 2013). Also consistent is the fact that each of the six patients whom we have characterized as carriers of the MITF E318K polymorphism had suffered from multiple primary melanomas and, apart from 149DC, relatively high nevus counts (Yokoyama et al., 2011). There was a high incidence of melanoma in situ and only one thick melanoma of 9 mm, a finding that does not support the suggestion that the 145

6 RA Sturm et al. a MITF E318K b MITF E318K c MITF WT Figure 2. Microphthalmia transcription factor (MITF)-staining patterns of melanomas from E318K mutation and wild-type (WT) carriers. The panels are representative staining patterns of melanomas obtained from (a, b) MITF E318K mutant and (c) WT patients. (a) An amelanotic melanoma in situ taken from the upper back of patient 149DC. (b) An invasive melanoma (Breslow thickness 1.1 mm) from the shoulder of M004RR. (c) A melanoma in situ respective to an invasive melanoma (Breslow thickness 0.42 mm) from the left upper back of a 43-year-old male MITF WT patient, 17DC. There is intense staining of melanocytic nuclei in each of the sections with no variation in the level of staining or pattern dependent upon MITF mutation status; however, there is an increased signal apparent when the cells form nests. Scale bars ¼ 100 (a), 200 (b), and 300 mm (c). MITF E318K mutation is associated with the development of a nodular subtype of melanoma (Ghiorzo et al., 2013); rather, it is more in agreement with there being no association with 146 Journal of Investigative Dermatology (2014), Volume 134 Breslow thickness as reported in the Australian/UK study (Yokoyama et al., 2011). It is possible that greater surveillance of those at high risk for melanoma in Australia could explain these differences as these cases of melanoma are more likely to be picked up early before progression to a nodular stage (Murali et al., 2012). Whole-body clinical examination of melanocytic nevi found on five MITF E318K patients did not distinguish any commonality of pigmentation phenotype, apart from fair skin. Dermoscopic analysis revealed that the predominant nevus signature pattern in nevi 45 þ 2 mm on the back was of a reticular type in both MITF E318K and WT patients, which was also the subtype with the greatest increase in number when the two groups were compared (Table 2). There was a high incidence of amelanotic melanomas found within the group, with 4 of 13 excised lesions presenting as nonpigmented melanoma; as 3 of the 5 melanomas from patient 149DC were amelanotic, it is possible that a genetic interaction between the MITF E318K allele and his MC1R red hair color (RHC) homozygous R/R genotype is the basis of this phenotype. Unfortunately, the pathology reports detailing the pigmentation status of the five melanoma lesions for patient 176SP, who was also of this MC1R genotype, were not available to confirm this observation. Previous association of the MC1R RHC alleles with amelanotic melanoma and depigmented nevi lends support to this interaction (Zalaudek et al., 2009; Curchin et al., 2012), and possibly for the high number of multiple primaries. Moreover, a genetic interaction has earlier been described for the high-penetrance familial melanoma gene CDKN2A and co-carriers of homozygous MC1R RHC alleles for the increasing incidence of multiple primary melanomas (Peris et al., 2004; Goldstein et al., 2005; Eliason et al., 2007; Pastorino et al., 2008). A higher frequency of MC1R RHC alleles has also been reported in CDK4 gene mutation melanoma prone family members with multiple primary melanomas (Puntervoll et al., 2013). A report on the dermoscopic features of primary melanomas in nine high-risk CDKN2A mutation carrier patients from the Spanish population found that the mean ABCD total dermoscopy score was significantly higher in those with MC1R non-rhc genotype (Cuellar et al., 2009) compared with those who carried two MC1R RHC variant alleles. Although the melanomas in CDKN2A patients homozygous for MC1R RHC alleles were more difficult to diagnose, with less color and fewer structures, atypical vessels were observed. The total dermoscopy score has been further examined in a larger group of 876 dysplastic nevi and 21 melanomas taken from high-risk CDKN2A European melanoma patients (Quint et al., 2012). Here it was reported that a pigment network, dark brown color, and streaks were more common in those without an MC1R RHC allele, although in contrast to the earlier report there was no difference in the total dermoscopy score of those who carried two MC1R RHC alleles. The dermoscopic and histopathological patterns of nevi associated with BRAF mutations have been recently reported (Zalaudek et al., 2011). Although the sensitivity of detection of BRAF changes among nevi was highly dependent on the detection method used, possibly because of clonal heterogeneity within the cells

7 of the nevus, BRAF mutations were more frequent in globular nevi, and in mixed reticular/homogeneous nevi with peripheral globules possibly representing growing nevi. The frequency of globular nevi in the five MITF E318K carriers in this report varied between patients; although the mean frequency was greater than MITF WT patients, this increase was not statistically significant (Table 2), and does not suggest that this mutation will be acting to force the continuous growth of nevi; more likely, the somatic mutations in BRAF within the nevi must occur to drive growth. It is clear from genetic and biochemical studies that the MITF protein acts as a transcription factor that regulates a suite of genes involved in pigment synthesis and cell cycle control (Cheli et al., 2010). As this is partly achieved through the MITF target gene APEX1 encoding APE1 and CDK2, we chose to perform IHC with antibodies to these proteins. Our finding of no statistical difference in staining patterns between MITF mutation-positive and -negative sections supports the conclusion that protein stability and nucleocytoplasmic partitioning are not affected by the E318K-induced SUMOylation deficiency of the MITF protein (Bertolotto et al., 2011; Yokoyama et al., 2011); rather, some differential regulation of gene expression will explain the increased risk for melanomagenesis caused by this mutation. It is clear that this may be a subset of MITF targets, but as the APE1 and CDK2 proteins did not differ in staining characteristics between mutation-positive and -negative sections the pathways concerned will have to be further characterized. MITFdirected gene regulation in melanocytic cell responses to reactive oxygen species stress (Liu et al., 2009), DNA damage (Beuret et al., 2011; Strub et al., 2011), the hypoxic response (Feige et al., 2011; Cheli et al., 2012), and melanoma initiator cell behavior and phenotype switching (Cheli et al., 2011; Thurber et al., 2011) are obvious candidates that must now be characterized in MITF E318K carrier patients. MATERIALS AND METHODS Study samples and DNA extraction Consecutive high-risk melanoma patients seen between October 2009 and October 2012, defined as those with a personal or a firstdegree family member with a history of melanoma, were recruited from the Melanoma Unit and Dermatology Department of the Princess Alexandra Hospital (PAH) or from Private Practice (HPS), which treats clients from the Southeast Queensland region. Moderateand low-risk melanoma volunteers were recruited through advertisement of our study throughout the PAH during 2010 and by direct contact through letters to participants in the Brisbane Twin Nevus Study (BTNS) requesting they volunteer (Duffy et al., 2010a). There were 301 individuals who provided saliva samples using an Oragene- DNA self-collection kit (DNA Genotec, Ottawa, ON, Canada) for genomic DNA processing. Every individual gave written informed consent; the study was conducted according to the Declaration of Helsinki Principles, and the University of Queensland HREC and QIMR HREC approved the ascertainment of samples. Measures and dermoscopy Of the 301 participants, 288 were able to undergo a detailed wholebody examination including digital recording of images from 16 body sites and individual dermoscopic images of all significant nevi recorded using a system for sequential total-body photography and dermoscopy (FotoFinder Systems GmbH, Bad Birnbach, Germany). Significant nevi were classified as those 42 mm on the back of both male and female participants and on the lower limbs of female participants, and those 45 mm at all other body sites (excluding scalp, buttocks, mucosal surfaces, and genitalia of both sexes). All significant nevi were classified in terms of their predominant dermoscopic pattern contributing to at least one-third of the lesions surface area (globular; reticular; globular/structureless with reticular rim; reticular/structureless with globular rim; and nonspecific, which includes homogenous nevi). Dermoscopically classified nevi were then grouped by age bracket (10 39 years and 40 þ years) and body site (head/neck, abdomen/chest, back, upper limbs, and lower limbs) for high- and moderate/low-risk melanoma groups. The signature nevus pattern (defined as the pattern seen in 440% nevi) (Scope et al., 2006) was calculated for each individual. A clinical assessment using standardized pigmentation characteristics was performed: hair color rated on a 5-category scale (1, fair/ blonde; 2, light brown; 3, red/auburn; 4, dark brown; and 5, black); eye color (1, blue/gray; 2, green/hazel; and 3, brown, which included standardized close-up digital imaging of the eyes to assess pigmentation pattern); skin color (inner upper left arm, innate/ facultative skin color (visually rated as 1, fair/pale; 2, medium; and 3, olive/dark, and also analyzed by spectrophotometry of the volar and dorsal skin of the arm) (CIE (L*, a*, b*) color space). Freckling density was recorded on a four-point scale (0, none; 1, mild/ infrequent/sparse; 2, moderate/evenly distributed; and 3, severe) at three body sites (face, dorsum of the right hand, and shoulders), with a composite freckling score constructed by summing the scores for these three sites (Duffy et al., 2004). DNA sequencing and genotyping Genomic DNA was subjected to automated DNA sequence analysis of PCR-amplified fragments covering the MC1R coding region (Box et al., 2001), CDKN2A exon 1a and exon 2 (Harland et al., 2008) using ABI PRISM Big Dye Terminator Sequencing (Applied Biosystems, Foster City, CA), with reactions processed by the Australian Equine Genetics Research Centre (AEGRC, Brisbane, Australia). Singlenucleotide polymorphism genotyping was performed using iplex Gold chemistry on a MALDI-TOF Mass Spectrometer (Sequenom, San Diego, CA) (Duffy et al., 2010b). TaqMan single-nucleotide polymorphism genotyping assays (Cook et al., 2009) were performed in a 96- or 384-well-plate format using a 7500 real-time PCR system and analyzed using 7500 Software (Applied Biosystems). Genotyping of rs MITF-E318K was carried out with Custom-designed Taqman assay AHWR8XV (Applied Biosystems). IHC of nevi and melanoma tumor samples A series of 32 lesions and pathology reports were obtained from the PA Hospital pathology archives and a further 7 lesions were collected from Sullivan and Nicolaides Pathology in Brisbane, Australia. Slides were deparaffinized and rehydrated through xylene to water. Antigen retrieval was performed with a 5-min boiling cycle in 0.01 M citrate buffer in a commercial decloaker. Peroxidase block was performed with 3% hydrogen peroxide for 15 minutes. Nonspecific binding was blocked with donkey serum/bsa in Tris-buffered saline for 30 minutes before overnight incubation with the primary antibody. The primary 147

8 antibodies used were monoclonal mouse anti-human APE1 (ab194, 1:10,000; AbCam, Cambridge, MA), monoclonal mouse anti-human Ki67 (M7240, 1:100; Dako, Campbellfield, VIC, Australia), monoclonal mouse anti-human MITF (M3621, 1:30; Dako), and polyclonal rabbit anti-human CDK2 (sc163 M2, 1:200; Santa Cruz Biotechnology, Santa Cruz, CA). Secondary antibody was Vector IMMPRESS (Burlingame, CA) anti-mouse or anti-rabbit. Vector ImmPACT NovaRED peroxidase substrate was applied as a chromogen for visualization of the immunoreaction. Hematoxylin was used for counterstaining. Negative controls were obtained by omitting the primary antibody. Slides were mounted using ClearMount mounting solution (Invitrogen, Mulgrave, VIC, Australia) and coverslipped with Xylene and Entellan (Merck Millipore, Kilsyth, VIC, Australia) mounting media. They were then scanned using an Aperio (Vista, CA) slide scanner at 20 magnification. CONFLICT OF INTEREST The authors state no conflict of interest. ACKNOWLEDGMENTS We thank Dr Duncan Lambie for help with pathology samples. RAS, DLD, and BG are Senior Research Fellows of the Australian NHMRC, and HPS has an NHMRC Practitioner Fellowship. MI-V is funded by the Spanish Ministerio de Educacion y Ciencia under a grant FPI (BES ). GR is funded by FI by the Spanish Ministerio de Salud. This work was funded by project grant NHMRC APP SUPPLEMENTARY MATERIAL Supplementary material is linked to the online version of the paper at REFERENCES Baxter AJ, Hughes MC, Kvaskoff M et al. (2008) The Queensland Study of Melanoma: environmental and genetic associations (Q-MEGA); study design, baseline characteristics, and repeatability of phenotype and sun exposure measures. Twin Res Hum 11: Beaumont KA, Wong SS, Ainger SA et al. (2011) Melanocortin MC(1) receptor in human genetics and model systems. Eur J Pharmacol 660: Bertolotto C, Lesueur F, Giuliano S et al. (2011) A SUMOylation-defective MITF germline mutation predisposes to melanoma and renal carcinoma. Nature 480:94 8 Beuret L, Ohanna M, Strub T et al. (2011) BRCA1 is a new MITF target gene. Pigment Cell Melanoma Res 24:725 7 Box NF, Duffy DL, Chen W et al. (2001) MC1R genotype modifies risk of melanoma in families segregating CDKN2A mutations. Am J Hum Genet 69: Cheli Y, Giuliano S, Botton T et al. (2011) Mitf is the key molecular switch between mouse or human melanoma initiating cells and their differentiated progeny. Oncogene 30: Cheli Y, Giuliano S, Fenouille N et al. (2012) Hypoxia and MITF control metastatic behaviour in mouse and human melanoma cells. Oncogene 31: Cheli Y, Ohanna M, Ballotti R et al. (2010) Fifteen-year quest for microphthalmia-associated transcription factor target genes. Pigment Cell Melanoma Res 23:27 40 Cook AL, Chen W, Thurber AE et al. (2009) Analysis of cultured human melanocytes based on polymorphisms within the SLC45A2/MATP, SLC24A5/NCKX5, and OCA2/P loci. J Invest Dermatol 129: Cronin JC, Wunderlich J, Loftus SK et al. (2009) Frequent mutations in the MITF pathway in melanoma. Pigment Cell Melanoma Res 22: Cuellar F, Puig S, Kolm I et al. (2009) Dermoscopic features of melanomas associated with MC1R variants in Spanish CDKN2A mutation carriers. Br J Dermatol 160:48 53 Curchin C, Wurm E, Jagirdar K et al. (2012) Dermoscopy, reflectance confocal microscopy and histopathology of an amelanotic melanoma from an individual heterozygous for MC1R and tyrosinase variant alleles. Australas J Dermatol 53:291 4 Douglas NC, Borgovan T, Carroll MJ et al. (2011) Dermoscopic naevus patterns in people at high versus moderate/low melanoma risk in Queensland. Australas J Dermatol 52: Du J, Widlund HR, Horstmann MA et al. (2004) Critical role of CDK2 for melanoma growth linked to its melanocyte-specific transcriptional regulation by MITF. Cancer Cell 6: Duffy DL, Box NF, Chen W et al. (2004) Interactive effects of MC1R and OCA2 on melanoma risk phenotypes. Hum Mol Genet 13: Duffy DL, Iles MM, Glass D et al. (2010a) IRF4 variants have age-specific effects on nevus count and predispose to melanoma. Am J Hum Genet 87:6 16 Duffy DL, Zhao ZZ, Sturm RA et al. (2010b) Multiple pigmentation gene polymorphisms account for a substantial proportion of risk of cutaneous malignant melanoma. J Invest Dermatol 130:520 8 Eliason MJ, Hansen CB, Hart M et al. (2007) Multiple primary melanomas in a CDKN2A mutation carrier exposed to ionizing radiation. Arch Dermatol 143: Feige E, Yokoyama S, Levy C et al. (2011) Hypoxia-induced transcriptional repression of the melanoma-associated oncogene MITF. Proc Natl Acad Sci USA 108:E Garraway LA, Widlund HR, Rubin MA et al. (2005) Integrative genomic analyses identify MITF as a lineage survival oncogene amplified in malignant melanoma. Nature 436: Ghiorzo P, Pastorino L, Queirolo P et al. (2013) Prevalence of the E318K MITF germline mutation in Italian melanoma patients: associations with histological subtypes and family cancer history. Pigment Cell Melanoma Res 26: Goldstein AM, Landi MT, Tsang S et al. (2005) Association of MC1R variants and risk of melanoma in melanoma-prone families with CDKN2A mutations. Cancer Epidemiol Biomarkers Prev 14: Granter SR, Weilbaecher KN, Quigley C et al. (2002) Role for microphthalmia transcription factor in the diagnosis of metastatic malignant melanoma. Appl Immunohistochem Mol Morphol 10:47 51 Harland M, Goldstein AM, Kukalizch K et al. (2008) A comparison of CDKN2A mutation detection within the Melanoma Genetics Consortium (GenoMEL). Eur J Cancer 44: Jaimes N, Braun RP, Thomas L et al. (2012) Clinical and dermoscopic characteristics of amelanotic melanomas that are not of the nodular subtype. J Eur Acad Dermatol Venereol 26:591 6 Law MH, Macgregor S, Hayward NK (2012) Melanoma genetics: recent findings take us beyond well-traveled pathways. JInvestDermatol132: Liu F, Fu Y, Meyskens FL Jr (2009) MiTF regulates cellular response to reactive oxygen species through transcriptional regulation of APE-1/Ref-1. J Invest Dermatol 129: McClain SE, Mayo KB, Shada AL et al. (2012) Amelanotic melanomas presenting as red skin lesions: a diagnostic challenge with potentially lethal consequences. Int J Dermatol 51:420 6 McClenahan P, Blake T, Douglas N et al. (2012) Quantifying the orientation of acquired melanocytic nevi on the back. Arch Dermatol 148:857 9 Murakami H, Arnheiter H (2005) Sumoylation modulates transcriptional activity of MITF in a promoter-specific manner. Pigment Cell Res 18: Murali R, Goumas C, Kricker A et al. (2012) Clinicopathologic features of incident and subsequent tumors in patients with multiple primary cutaneous melanomas. Ann Surg Oncol 19: Pastorino L, Bonelli L, Ghiorzo P et al. (2008) CDKN2A mutations and MC1R variants in Italian patients with single or multiple primary melanoma. Pigment Cell Melanoma Res 21:700 9 Peris K, Fargnoli MC, Pacifico A et al. (2004) CDKN2A and MC1R mutations in patients with sporadic multiple primary melanoma. J Invest Dermatol 122: Puntervoll HE, Yang XR, Vetti HH et al. (2013) Melanoma prone families with CDK4 germline mutation: phenotypic profile and associations with MC1R variants. J Med Genet 50: Journal of Investigative Dermatology (2014), Volume 134

9 Quint KD, van der Rhee JI, Gruis NA et al. (2012) Melanocortin 1 receptor (MC1R) variants in high melanoma risk patients are associated with specific dermoscopic ABCD features. Acta Derm Venereol 92: Salama AK, de Rosa N, Scheri RP et al. (2013) Hazard-rate analysis and patterns of recurrence in early stage melanoma: moving towards a rationally designed surveillance strategy. PloS One 8:e57665 Scope A, Burroni M, Agero AL et al. (2006) Predominant dermoscopic patterns observed among nevi. J Cutan Med Surg 10:170 4 Strub T, Giuliano S, Ye T et al. (2011) Essential role of microphthalmia transcription factor for DNA replication, mitosis and genomic stability in melanoma. Oncogene 30: Thurber AE, Douglas G, Sturm EC et al. (2011) Inverse expression states of the BRN2 and MITF transcription factors in melanoma spheres and tumour xenografts regulate the NOTCH pathway. Oncogene 30: Yokoyama S, Woods SL, Boyle GM et al. (2011) A novel recurrent mutation in MITF predisposes to familial and sporadic melanoma. Nature 480: Zalaudek I, Guelly C, Pellacani G et al. (2011) The dermoscopical and histopathological patterns of nevi correlate with the frequency of BRAF mutations. J Invest Dermat 131:542 5 Zalaudek I, Meiklejohn W, Argenziano G et al. (2009) White nevi and red melanomas: association with the RHC phenotype of the MC1R gene. J Invest Dermat 129: