Specific mosaic KRAS mutations affecting codon 146 cause oculoectodermal syndrome and encephalocraniocutaneous lipomatosis

Transcription

1 Clin Genet 2016: 90: Printed in Singapore. All rights reserved Original Article 2016 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd CLINICAL GENETICS doi: /cge Specific mosaic KRAS mutations affecting codon 146 cause oculoectodermal syndrome and encephalocraniocutaneous lipomatosis Boppudi S., Bögershausen N., Hove H.B., Percin E.F., Aslan D., Dvorsky R., Kayhan G., Li Y., Cursiefen C., Tantcheva-Poor I., Toft P.B., Bartsch O., Lissewski C., Wieland I., Jakubiczka S., Wollnik B., Ahmadian M.R., Heindl L.M., Zenker M. Specific mosaic KRAS mutations affecting codon 146 cause oculoectodermal syndrome and encephalocraniocutaneous lipomatosis. Clin Genet 2016: 90: John Wiley & Sons A/S. Published by John Wiley & Sons Ltd, 2016 Oculoectodermal syndrome (OES) and encephalocraniocutaneous lipomatosis (ECCL) are rare disorders that share many common features, such as epibulbar dermoids, aplasia cutis congenita, pigmentary changes following Blaschko lines, bony tumor-like lesions, and others. About 20 cases with OES and more than 50 patients with ECCL have been reported. Both diseases were proposed to represent mosaic disorders, but only very recently whole-genome sequencing has led to the identification of somatic KRAS mutations, p.leu19phe and p.gly13asp, in affected tissue from two individuals with OES. Here we report the results of molecular genetic studies in three patients with OES and one with ECCL. In all four cases, Sanger sequencing of the KRAS gene in DNA from lesional tissue detected mutations affecting codon 146 (p.ala146val, p.ala146thr) at variable levels of mosaicism. Our findings thus corroborate the evidence of OES being a mosaic RASopathy and confirm the common etiology of OES and ECCL. KRAS codon 146 mutations, as well as the previously reported OES-associated alterations, are known oncogenic KRAS mutations with distinct functional consequences. Considering the phenotype and genotype spectrum of mosaic RASopathies, these findings suggest that the wide phenotypic variability does not only depend on the tissue distribution but also on the specific genotype. Conflict of interest All authors declare that they have no conflict of interest. S. Boppudi a, N. Bögershausen b,c, H.B. Hove d,e.f.percin e, D. Aslan f,r.dvorsky g, G. Kayhan e, Y. Li b,c, C. Cursiefen h, I. Tantcheva-Poor i, P.B. Toft j, O. Bartsch k, C. Lissewski a, I. Wieland a, S. Jakubiczka a, B. Wollnik b,c, M.R. Ahmadian g, L.M. Heindl h, and M. Zenker a, a Institute of Human Genetics, University Hospital Magdeburg, Otto-von-Guericke University, Magdeburg, Germany, b Institute of Human Genetics, University Medical Center Goettingen, Georg-August University, Goettingen, Germany, c Institute of Human Genetics, University of Cologne, Cologne, Germany, d Department of Clinical Genetics, Copenhagen University Hospital Rigshospitalet, Copenhagen, Denmark, e Department of Medical Genetics, Faculty of Medicine, Gazi University, Ankara, Turkey, f Section of Hematology, Department of Pediatrics, Faculty of Medicine, Gazi University, Ankara, Turkey, g Institute of Biochemistry and Molecular Biology II, Medical Faculty, Heinrich-Heine-University, Düsseldorf, Germany, h Department of Ophthalmology, University of Cologne, Cologne, Germany, i Department of Dermatology, University of Cologne, Cologne, Germany, j Department of Ophthalmology, Copenhagen University Hospital Rigshospitalet, Copenhagen, Denmark, and k Institute of Human Genetics, University Medical Centre of the Johannes Gutenberg University Mainz, Mainz, Germany These authors contributed equally as senior authors 334

2 Specific mosaic KRAS mutations affecting codon 146 cause OES and ECCL Key words: encephalocraniocutaneous lipomatosis KRAS mosaic RASopathy oculoectodermal syndrome somatic mutation Corresponding author: Prof Dr med Martin Zenker, Institute of Human Genetics, University Hospital Magdeburg, Otto-von-Guericke University, Leipziger Str. 44, Magdeburg, Germany. Tel.: ; fax: ; Received 29 January 2016, revised and accepted for publication 6 March 2016 Oculoectodermal syndrome (OES, OMIM ) is a very rare disorder that was first described in 1993 by Toriello et al. in two unrelated individuals (1). About 20 cases have been reported in the literature [reviewed in Refs (2, 3)]. OES is characterized by the association of epibulbar dermoids and congenital scalp lesions referred to as aplasia cutis congenita (ACC). The latter represent atrophic, non-scarring, hairless regions, and they are often multiple and asymmetric in their distribution. Hamartomas may be associated with the regions of alopecia (myxovascular hamartoma, smooth muscle hamartoma) (2, 4). Furthermore, ectodermal changes include linear hyperpigmentation that may follow the lines of Blaschko and rarely epidermal nevus-like lesions. Epibulbar dermoids may be uni- or bilateral. Additional ocular anomalies such as skin tags of the upper eyelid, rarely optic nerve or retinal changes, and microphthalmia can be present. The phenotypic expression is highly variable, and various other abnormalities have occasionally been reported including growth failure, lymphedema, cardiovascular defects, as well as neurodevelopmental symptoms like developmental delay, epilepsy, learning difficulties, and behavioral abnormalities (2). Benign tumor-like lesions such as non-ossifying fibromas of the long bones and giant cell granulomas of the jaws have repeatedly been observed and appear to be age-dependent, becoming a common manifestation in individuals aged 5 years or older (3, 5). OES shares many common features with encephalocraniocutaneous lipomatosis (ECCL), and it has long been assumed that OES and ECCL belonged to the same spectrum of disorders that might be caused by mutations in the same gene (2, 6). The common manifestations in both disorders include ACC/focal alopecia, epibulbar dermoids, upper eyelid lesions, and linear hyperpigmentation. In addition, giant cell granulomas of the jaws and non-ossifying fibromas of long bones have also been reported in ECCL (6). A distinct hairless fatty tissue nevus of the scalp (nevus psiloliparus) is regarded as the dermatological hallmark of ECCL (7). Subcutaneous fatty masses in the frontotemporal or zygomatic region are common in ECCL, but have occasionally been reported also in children diagnosed with OES. Central nervous system involvement is a distinguishing feature of ECCL and includes intracranial and intraspinal lipomas, arachnoid cysts, cerebral defects of putative vascular origin, developmental retardation, and seizures (8). In both, OES and ECCL, exclusively sporadic occurrence has been observed. Together with the obvious mosaic pattern of skin involvement, this is suggestive of genomic mosaicism with mutations that would confer embryonic lethality when occurring in the germline. Recently, Peacock et al. identified mutations in the KRAS gene (V-Ki-Ras2 Kirsten rat sarcoma viral oncogene homolog; OMIM ), namely c.38g>a (p.gly13asp) and c.57g>c (p.leu19phe), in affected tissues from two patients with OES, thus suggesting that OES is a mosaic RASopathy (3). Here, we present three further patients with OES and one with ECCL in all of which mosaic mutations in the KRAS gene could be shown in lesional tissue. Clinical reports Patient 1 The propositus is the first child of healthy nonconsanguineous parents of German origin. He was born at term after an uneventful pregnancy via normal spontaneous delivery. His birth weight was 3100 g ( 0.6 SD), his length was 51 cm ( 0.2 SD), and his head circumference 34 cm ( 0.4 SD). He did not experience perinatal adaptation problems. At birth, small skin tags of the right upper eyelid and the right eyebrow as well as an epibulbar dermoid on the right ocular surface, involving conjunctiva as well as cornea, were noted (Fig. 1a,b). The skin tags measured approximately 3 4 mm and were excised at age 4.5 years. Histopathology confirmed the lesions to be fibroepitheliomas. The epibulbar tumor was excised at age 4.7 years, and histopathology showed mature tissue containing compounds of cartilage and skin, in line with the clinical diagnosis of a dermoid. At clinical genetic evaluation he presented as an active, friendly boy with normal psychomotor development. He had a triangular face with prominent zygomatic arches and mild frontal bossing, a frontal upsweep, as well as 335

3 Boppudi et al. (a) (c) (a) (b) (c) (b) (d) (d) (e) (f) Fig. 1. Patient 1. (a, b) Right epibulbar dermoid, eyelid lesion and small skin tag of the eyebrow. (c) Parietal region of alopecia. (d) Histology from this region showing a linear arrangement of the arrector pili muscles in the midcorium (black arrows, HE-staining). deeply set eyes with a relatively narrow intercanthal distance. He had a hairless patch with immovable, yellowish, leathery skin on his right frontoparietal scalp, which measured approximately 8 4 cm (Fig. 1c). Histological examination of a biopsy of this lesion revealed a paucity of hair follicles without scarring and loss of elastic fibers in the trichrome-masson staining. It was noticeable that the arrector pili muscles were arranged in the midcorium in a line parallel to the surface (Fig. 1d). We could not properly evaluate the subcutis because the biopsy was rather superficial a protrusion of the fat tissue into the reticular dermis was however suspected. Based on these features, histopathology was interpreted as a hamartoma with features suggestive of nevus psiloliparus, which is generally regarded as the cutaneous hallmark of the ECCL but has also been reported in OES (2). Because of the presentation of the three cardinal features: fibroma of the eyelids/eyebrows, epibulbar dermoid and the typical scalp lesion in a boy with normal psychomotor development, a clinical diagnosis of OES was made. A blood sample and a native tissue sample from the resected epibulbar dermoid were available for DNA extraction and genetic analysis. Patient 2 This male patient was born at term with normal birth parameters and a relatively large head circumference of 37 cm (+1.5 SD). The parents are non-consanguineous, of Danish descent, and the patient is their first child. Fig. 2. Patient 2. (a c) Clinical photos at the age of 10 months showing bilateral ocular dermoids, skin tags, hypoplasia of upper eyelids, and scalp lesions. (d, e) The same patient at age 5 years after ocular surgery. Note focal alopecia of the scalp. (f) X-ray image of the left upper limb with cystic lesions in the humerus and proximal ulna. He presented with bilateral epibulbar dermoids, everted and hypoplastic upper eyelids with skin tags, and four well-circumscribed alopecic lesions on the upper right parietal scalp (Fig. 2). Eyelid defects required plastic surgery at age 1 year, and strabismus was corrected at age 1.5 years. At age 5 years, binocular vision of 6/24 was measured using symbol charts (Osterberg symbols), indicating moderate visual impairment. Magnetic resonance imaging (MRI) of the brain performed at age 9 months showed hemiatrophy of left cerebral hemisphere and ventriculomegaly, but no definite evidence of intracranial lipomatosis. The boy had normal cognitive development, while his motor development was mildly delayed, probably because of visual impairment. He walked at years and spoke his first words at 11 months. He has had no seizures or other neurological symptoms. At age 4 years, aortic coarctation with a gradient of 34 mm Hg was diagnosed by echocardiography. A follow-up computed tomography (CT) scan at age 4.5 years additionally revealed a narrow dimension of the left main bronchus causing a partial atelectasis, and a cystic paravertebral process was identified at the level of Th10/Th11. Cyst puncture did not yield any malignant cells. A skeletal survey showed several benign cystic lesions in the long bones of all four extremities and clavicles (Fig. 2f). At the last clinical evaluation at the age 5 years, he showed mild growth failure (height: cm, 2.4 SD; weight: 16.0 kg, 1.5 SD) with relative macrocephaly (head circumference: 53.5 cm, +1.5 SD). He had focal alopecia, streaks of pigmentary changes on the neck and both arms, and atopic dermatitis 336

4 Specific mosaic KRAS mutations affecting codon 146 cause OES and ECCL in several locations on the extremities and cheeks. No limb asymmetry was noted. Karyotype was normal, and array comparative genomic hybridisation (CGH) at 244 k resolution showed no copy number variation. Based on these multiple abnormalities, the clinical diagnosis of ECCL was made according to published diagnostic criteria (6). A blood sample and a native skin biopsy from the scalp lesion were available for genetic analysis. Patient 3 The clinical phenotype of this male 3.5 year old patient was recently described (9). He is the first child of healthy non-consanguineous Turkish parents. His characteristic features supporting the diagnosis of OES included hairless areas in the parieto-occipital region of the scalp consistent with ACC, a left-sided epibulbar dermoid, multiple areas of pigmentary changes following the lines of Blaschko mainly on the neck and trunk, a tumor of the lower jaw histologically classified as giant cell granuloma, and multiple cystic lesions of the right humerus and clavicle suggesting non-ossifying fibromas. Body length, weight and head circumference were in the normal ranges. A blood sample as well as fibroblasts from a skin biopsy and paraffin-embedded tumor tissue from a giant cell granuloma of the mandible were available for DNA extraction and genetic analysis. Patient 4 The clinical phenotype of this 6 year old male was reported by Aslan et al. as a case of OES (10). His clinical abnormalities included bilateral epibulbar dermoids, multiple areas of scalp alopecia, streaky areas of hypoand hyperpigmentation following the lines of Blaschko, intellectual disability, and attention deficit hyperactivity disorder. Brain MRI showed enlarged cisterna magna, an enlarged fluid space in the quadrigeminal cistern suggesting an arachnoid cyst, and several subcutaneous masses within the scalp. His anthropometric measurements were normal. Formally, he would also fulfill the criteria for a probable diagnosis of ECCL according to Moog et al. (6) but classification as OES was preferred, as there was no intracranial or intraspinal lipoma. A blood sample as well as biopsies from a scalp lesion and a hyperpigmented area of the skin were available for genetic analysis. Methods Informed consent was obtained for the collection of all blood and tissue samples for genetic testing. For patients 1 and 3, remaining tissue samples from surgical removal of epibulbar dermoids, and a mandibular giant cell granuloma, respectively, were used for the analysis. Patients 2, 3 and 4 had small biopsies of the scalp lesions and from hyperpigmented areas of the skin, respectively, to obtain material for genetic testing. The skin biopsy samples from patients 2 and 4 were divided into a superficial part containing mainly the epidermal layer and a deeper fraction containing dermis and part of the subcutaneous layer. In addition, a cell culture was developed from skin fibroblasts of patient 2. For patient 3, only the fibroblast line but no native tissue was available from the skin biopsy. Parental written permission was obtained to publish the patients photographs. Genomic DNA was isolated from peripheral leukocytes, tissue samples, and cell cultures according to standard procedures. All coding exons with flanking introns of KRAS gene were amplified by polymerase chain reaction, and bidirectional Sanger sequencing was performed using Big Dye Terminator Cycle Sequencing Kit and a 3500xl Genetic Analyzer (Applied Biosystems, Foster City, CA). Sequences were aligned using the Seqpilot analysis software (JSI medical systems, Kippenheim, Germany) and compared with the reference sequences (KRASB; ENST ). Mosaic level for mutations was estimated by comparing the area under the curve of electropherograms for the wildtype and mutant peaks in the forward and reverse sequencing directions using the Seqpilot software. Results In lesional tissues from all four patients, we could identify KRAS mutations at various levels of mosaicism, while the mutations were not detected in leukocyte-derived DNA at the detection threshold of Sanger sequencing. The observed mutations consistently affected exon 4, codon 146. Two patients were found to carry a mosaic for the sequence variant c.437c>t (p.ala146val), and in the other two, the mutation c.436g>a (p.ala146thr) was identified (Table 1; Fig. S1, Supporting information). The calculated proportion of the mutated allele ranged from 11% to nearly 50% in the dermal fibroblast cultures from patient 2, indicating a non-mosaic heterozygous pattern in these cells (Fig. S1b). We were unable to robustly confirm the mutation in the DNA extracted from a formalin-fixed paraffin-embedded tissue sample derived from the giant cell lesion of the mandible of patient 3, but this could be because of a very low amount and quality of DNA that prevented reliable sequencing results (Fig. S1c). Both the observed alterations at KRAS codon 146 are known cancer-associated mutations and listed in the COSMIC database (cancer.sanger.ac.uk/cosmic; mutation IDs: COSM19404, COSM19900). They are consistently predicted to be pathogenic by various web-based prediction tools (Mutation Taster, taster.org; Polyphen-2, genetics.bwh.harvard.edu/pph2; UMD-Predictor, umd-predictor.eu; PhD-SNP, snps.bio fold.org/phd-snp). Discussion Recently, Peacock et al. first reported mosaic KRAS mutations, p.gly13asp and p.leu19phe, in two OES patients, suggesting that OES is a mosaic RASopathy (3). The findings presented here corroborate that OES 337

5 Boppudi et al. Table 1. Results of KRAS genotyping in various tissues Tissue sample(s): Patient # KRAS mutation Type (proportion of mutant allele a ) Leukocyte DNA 1 c.437c>t (p.ala146val) Epibulbar dermoid (17%) Negative b 2 c.436g>a (p.ala146thr) Skin biopsy of scalp lesion: Negative b Superficial (epidermal) fraction (43%) Lower (dermal) fraction (35%) Cultivated fibroblasts (50%) 3 c.437c>t (p.ala146val) Fibroblasts from skin biopsy (40%) Negative b Jaw giant cell tumor (<10%) 4 c.436g>a (p.ala146thr) Skin biopsy of scalp lesion: Negative b Superficial (epidermal) fraction (24%) Lower (dermal) fraction (38%) Skin biopsy of hyperpigmented skin: Superficial (epidermal) fraction (22%) Lower (dermal) fraction (11%) a A proportion of the mutant allele of 50% represents a non-mosaic pattern for a heterozygous mutation. b Below detection threshold of Sanger sequencing. consistently results from a mosaic status for specific KRAS mutations and confirm for the first time the common genetic etiology of ECCL, which was the clinical diagnosis in patient 2. Both OES and ECCL, thus, belong to the mosaic RASopathies, a rapidly growing group of neurocutaneous disorders. The term RASopathies has originally been coined for a group of diseases comprising neurofibromatosis type 1 as well as Noonan syndrome and related disorders such as cardiofaciocutaneous, and Costello syndromes, which are all caused by germline mutations in genes encoding various components or regulators of the RAS-mitogen-activated protein kinase (MAPK) signaling pathway (11, 12). Short stature, cardiovascular and lymphatic abnormalities, a characteristic craniofacial phenotype, variable intellectual disability, and a predisposition for tumor development are the hallmarks of this group of disorders. The RASopathy-associated mutations lead to a dysregulation (mostly over-activation) of RAS-MAPK signaling, similar to somatic mutations that can be found in the same genes as oncogenic drivers of tumorigenesis (13). However, there is a strong body of evidence that RASopathy-associated germline mutations in KRAS and other RAS-MAPK pathway genes are not identical with their oncogenic counterparts, and that they rather represent hypomorphs leading to significantly milder levels of over-activation than oncogenic mutations that are not tolerated in the germline (14). HRAS mutations are an exception from this rule, as typical oncogenic changes of this gene are not lethal in the germline but do occur in Costello syndrome. Notably, also HRAS c.436g>a (p.ala146thr) and c.437c>t (p.ala146val), the mutations correspondingtothekras alterations we observed in OES/ECCL, have been reported in Costello syndrome (15, 16). Patients with mosaic mutations can occasionally be observed in neurofibromatosis type 1, and two cases of Costello syndrome with mosaic HRAS mutations have been described (17, 18). Similarly, mosaic cases may also exist for other RASopathies. Such mosaic manifestations belong to the categories of type 1 and 2 segmental mosaicism of autosomal dominant disorders according to a recently proposed classification (19). The term mosaic RASopathies has been introduced and is now mainly used for a group of (neuro)cutaneous disorders where typically the oncogenic type of mutations can be found in affected tissues (20). They thus represent manifestations of lethal mutations surviving by mosaicism according to the aforementioned classification (19). The predominant description of mosaic RASopathies as disorders with ectodermal involvement may be because the skin is the part of the body where mosaicism is most easily detected, and we still may not be aware of the full spectrum of anomalies that may be caused by mosaic RAS-MAPK pathway mutations. Clinically, mosaic RASopathies appear to have little in common with the germline RASopathies, and Noonan syndrome-like features are usually not recognizable. This may be related to the restricted tissue distribution of the mosaic mutations as well as to the more severe effects in mutated cells and tissues. Notably however, in OES and ECCL, we noticed that among the less commonly reported abnormalities, there are also some typical symptoms that occur in germline RASopathies, such as congenital heart defects (atrial septal defects, aortic coarctation), fetal hydrothorax, lymphedema, macrocephaly, large birth measurements, and relative or absolute short stature (1, 2, 6). Moreover, specific tumor types have repeatedly been reported in both, germline RASopathies as well as OES/ECCL, such as giant cell tumors and embryonal rhabdomyosarcoma (21 23). Oncogenic KRAS mutations, now established as the genetic basis of OES/ECCL, provide a reasonable explanation for such tumor associations, and consequently regular clinical follow-up should be recommended for early diagnosis of oncologic complications. Mosaic RASopathies constitute a rapidly expanding group of disorders. The spectrum ranges from simple nevoid skin lesions to their systematized forms and multiorgan involvement as exemplified with isolated nevus 338

6 Table 2. Mosaic RASopathies and associated mutations Specific mosaic KRAS mutations affecting codon 146 cause OES and ECCL Entity Gene Mutation(s) a Reference(s) Nevus sebaceus HRAS G13R (108) (35 37) G12S (3) G12D (2) G12C (1) A11S (1) KRAS G12D (5) G12V (2) Schimmelpenning / linear nevus sebaceous syndrome HRAS G13R (3) (35, 37 41) KRAS G12D (3) NRAS Q61R (1) Non-organoid keratinocytic epidermal nevus HRAS G13R (21) (42) G12C (1) G12V (1) Q61L (1) NRAS G12D (1) P34L (1) Q61R (1) KRAS G12D (1) Linear syringocystadenoma papilliferum syndrome BRAF V600E (1) (43) Phakomatosis pigmentokeratotica HRAS G13R (4) (44) Q61R (2) Nevus spilus HRAS G13R (8) (45) Oculoectodermal syndrome / ECCL KRAS G13D (1) (3); this report L19F (1) A146V (2) A146T (2) Neurocutaneous melanosis / Congenital giant melanocytic nevi NRAS Q61K (55) (46 48) Q61R (26) BRAF V600E (6) Nevus spilus-type congenital melanocytic nevi NRAS Q61H (2) (33, 34) Q61L (1) G13R (1) Wooly hair nevus HRAS G12S (2) (49) Cutaneous-skeletal hypophosphatemia syndrome (CSHS) HRAS G13R (2) (50) Q61R (1) NRAS Q61R (2) a Numbers in parentheses indicate the number of unrelated samples with the individual mutations. sebaceous and Schimmelpenning syndrome (20, 24) (Table 2). A characteristic of mosaic disorders is their extreme variability of phenotypic expression. The developmental timing when a mutation occurs and hence its distribution over the body are major determinants for this variation (25). The same applies for mosaic RASopathies, as, for example, the same recurrent HRAS mutation p.gly13arg predominates in different clinical manifestations, obviously depending just on its tissue distribution: nevus sebaceous and Schimmelpenning syndrome, non-organoid keratinocytic epidermal nevus, nevus spilus, and phakomatosis pigmentokeratotica (Table 2). The second important determining factor for the phenotype of mosaic RASopathies appears to be the genotype, as certain disorders are related predominantly or exclusively to specific genes. For example, congenital giant melanocytic nevi and neurocutaneous melanosis are associated with NRAS and BRAF mutations, while sebaceous nevi and Schimmelpenning syndrome are caused by HRAS and KRAS mutations (Table 2). Similar observations have been made with oncogenic mutations in cancer where alterations of particular RAS isoforms are enriched in specific types of tumors (26). But significant genotype correlations are not only regarding the different RAS isoforms but appear to exist also for specific mutations in the same gene. Mosaic KRAS mutations that have previously been observed in some cases of nevus sebaceous, Schimmelpenning syndrome, and in non-organoid keratinocytic nevi are p.gly12asp and p.gly12val (Table 2). In contrast, OES has previously been associated with the mutations p.leu19phe and p.gly13asp (3), and we found p.ala146val and p.ala146thr in OES/ECCL. Although the numbers of cases are still quite low, it appears unlikely that these non-overlapping mutation spectra in different types of KRAS mosaic diseases differ just by chance. It has been shown for germline as well as for somatic cancer-associated mutations that different KRAS alterations may have unique biochemical effects 339

7 Boppudi et al. (a) (b) Fig. 3. Protein structure and functional consequences of KRAS mutations. Relative positions of the three amino acids, Gly-13, Leu-19 and Ala-146 are illustrated in the structure of KRAS displayed as ribbon diagram with a bound GTP nucleotide shown as stick model (a); the same positions altered to aspartic acid (Asp), phenylalanine (Phe) and threonine (Thr), representing OES/ECCL-associated mutations (b). These three highly conserved amino acids are in the vicinity or part of the nucleotide binding pocket of KRAS. Gly-13 is located on the edge of nucleotide binding site. Introduction of a negatively charged residue like aspartic acid at this position is predicted to interfere with negatively charged phosphates (orange sticks with red ends) of GTP. Ala-146 and Leu-19 side chains are neighboring in the three-dimensional structure of KRAS and involved in shaping of the part of nucleotide binding pocket that harbors the guanine ring of GTP. Their substitutions by bulkier residues are predicted to alter nucleotide binding properties. The common consequences based on the illustrated sterical relationships for all three altered residues are likely accelerated GTP dissociation kinetics, in line with the findings of functional testing for some of these mutations (26, 28). (14, 26). RAS molecules act as molecular switches by cycling between an active guanosine triphosphate (GTP)-bound and an inactive guanosine diphosphate (GDP)-bound form (27). Guanine nucleotide exchange factors (GEFs) activate RAS by stimulating the release of GDP, thereby allowing binding of GTP which is abundant in the cytosol. On the other hand, GTPase-activating proteins (GAPs) terminate activation by stimulation of RAS GTPase activity. Activating effects of RAS mutations may be because of accelerated (GEF-independent) GDP/GTP exchange or defective GTPase activity (GAP resistance). The latter is the major activating mechanism of the most common cancer-associated KRAS mutations (such as Gly12Asp, Gly12Val and others) (26). Alterations of Gly-13, Leu-19 and Ala-146 are instead predicted to rather affect GDP/GTP exchange kinetics because of their specific spatial relationship to the nucleotide binding cleft (Fig. 3). Indeed, RAS Ala146Val and Gly13Asp mutants were both showed to have a strongly increased nucleotide exchange rate with no and only minor impairment of GTPase activity, respectively (26, 28). KRAS Leu19Phe has been shown to accumulate in the GTP-bound state and to have oncogenic activity (29). The precise activating mechanism for this mutant has not been resolved, but the position of Leu-19 relative to Ala-146 would be compatible with the assumption that functional effects might be similar (Fig. 3). Notably, in NIH3T3 cells expressing mutant KRAS isoforms, transcription-profiling experiments and hierarchical clustering analysis revealed the presence of two major clusters for differential gene expression, one of them containing Gly13Asp, Ala146Thr, and Leu19Phe, and one containing the Gly12Val, Gly12Cys and Gly12Asp mutants (30). Although it is too preliminary to speculate about the exact pathophysiological mechanisms, these findings suggest that all hitherto known OES/ECCL-associated mutations generate KRAS proteins with similarly altered biological properties probably related to their ability to over-activate through GEF-independent nucleotide exchange. These specific functional consequences may determine the phenotype because of distinct and tissue-dependent effects on cell fate decisions. This hypothesis is further supported by the observation that specific oncogenic RAS mutations also display differential coupling to specific cancers (31). KRAS p.gly13asp as well as codon 146 mutations are commonly found in colorectal cancer but infrequent in pancreatic or lung cancer (30, 32). Moreover, mutation-specific phenotype associations have also been observed in other mosaic RASopathies, as shown for NRAS mutations in congenital melanocytic nevi (CMN) vs the nevus spilus type of CMN (33, 34) (Table 2). Nevertheless, despite significant point mutation biases, the borders between mosaic RASopathy phenotypes and their association with specific mutations may not be absolute. In conclusion, we establish OES and ECCL as mosaic RASopathies and define codon 146 of KRAS as a hotspot for mutations associated with these related disorders. Despite some overlaps between the various mosaic RASopathies both phenotypically and genotypically, there is growing evidence for mutation-specific phenotype associations, the pathophysiological basis of which needs to be addressed by future research. 340

8 Specific mosaic KRAS mutations affecting codon 146 cause OES and ECCL Supporting Information Additional supporting information may be found in the online version of this article at the publisher s web-site. Acknowledgements The authors wish to express their gratitude to the patients and their families for participation in this study. We thank Dr R. Fikret Akata for supporting the collection of material from patient 4, and Susan Engelberg and Nicole Epperlein for excellent technical support. This study was funded by: German Research Foundation, DFG ZE 524/10-1; German Federal Ministry of Education and Research (BMBF, NSEuroNet and GeNeRARe); German Research Foundation, DFG Research Unit FOR2240 (HE 6743/2-1; HE 6743/3-1; RE 3777/1-1; CU47/6-1; CU47/12-1). Ethics approval All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards. For this type of study formal consent is not required. Informed consent Informed consent was obtained from all individual participants included in the study. Additional informed consent was obtained from all individual participants for whom identifying information is included in this article. References 1. Toriello HV, Lacassie Y, Droste P, Higgins JV. Provisionally unique syndrome of ocular and ectodermal defects in two unrelated boys. Am J Med Genet 1993: 45: Ardinger HH, Horii KA, Begleiter ML. Expanding the phenotype of oculoectodermal syndrome: possible relationship to encephalocraniocutaneous lipomatosis. Am J Med Genet A 2007: 143A: Peacock JD, Dykema KJ, Toriello HV et al. Oculoectodermal syndrome is a mosaic RASopathy associated with KRAS alterations. Am J Med Genet A 2015: 167: Gunduz K, Shields CL, Doych Y, Schnall B, Shields JA. Ocular ectodermal syndrome of epibulbar dermoid and cutaneous myxovascular hamartoma. Br J Ophthalmol 2000: 84: Federici S, Griffiths D, Siberchicot F, Chateil JF, Gilbert B, Lacombe D. Oculo-ectodermal syndrome: a new tumour predisposition syndrome. Clin Dysmorphol 2004: 13: Moog U. Encephalocraniocutaneous lipomatosis. J Med Genet 2009: 46: Happle R, Kuster W. Nevus psiloliparus: a distinct fatty tissue nevus. Dermatology 1998: 197: Moog U, Jones MC, Viskochil DH, Verloes A, Van Allen MI, Dobyns WB. Brain anomalies in encephalocraniocutaneous lipomatosis. Am J Med Genet A 2007: 143A: Mermer S, Kayhan G, Karacelebi E, Percin EF. Oculoectodermal syndrome: a new case with giant cell granulomas and non-ossifying fibromas. Genet Counsel 2015 (in press). 10. Aslan D, Akata RF, Schroder J, Happle R, Moog U, Bartsch O. Oculoectodermal syndrome: report of a new case with a broad clinical spectrum. Am J Med Genet A 2014: 164A: Tidyman WE, Rauen KA. The RASopathies: developmental syndromes of Ras/MAPK pathway dysregulation. Curr Opin Genet Dev 2009: 19: Zenker M. Clinical manifestations of mutations in RAS and related intracellular signal transduction factors. Curr Opin Pediatr 2011: 23: Schubbert S, Bollag G, Shannon K. Deregulated Ras signaling in developmental disorders: new tricks for an old dog. Curr Opin Genet Dev 2007: 17: Gremer L, Merbitz-Zahradnik T, Dvorsky R et al. Germline KRAS mutations cause aberrant biochemical and physical properties leading to developmental disorders. Hum Mutat 2011: 32: Gripp KW, Innes AM, Axelrad ME et al. Costello syndrome associated with novel germline HRAS mutations: an attenuated phenotype? Am J Med Genet A 2008: 146A: Zampino G, Pantaleoni F, Carta C et al. Diversity, parental germline origin, and phenotypic spectrum of de novo HRAS missense changes in Costello syndrome. Hum Mutat 2007: 28: Gripp KW, Stabley DL, Nicholson L, Hoffman JD, Sol-Church K. Somatic mosaicism for an HRAS mutation causes Costello syndrome. Am J Med Genet A 2006: 140: Sol-Church K, Stabley DL, Demmer LA et al. Male-to-male transmission of Costello syndrome: G12S HRAS germline mutation inherited from a father with somatic mosaicism. Am J Med Genet A 2009: 149A: Happle R. The categories of cutaneous mosaicism: a proposed classification. Am J Med Genet A 2016: 170: Luo S, Tsao H. Epidermal, sebaceous, and melanocytic nevoid proliferations are spectrums of mosaic RASopathies. J Invest Dermatol 2014: 134: Kratz CP, Rapisuwon S, Reed H et al. Cancer in Noonan, Costello, cardiofaciocutaneous and LEOPARD syndromes. Am J Med Genet C Semin Med Genet 2011: 157C: Lees M, Taylor D, Atherton D, Reardon W. Oculo-ectodermal syndrome: report of two further cases. Am J Med Genet 2000: 91: Neumann TE, Allanson J, Kavamura I et al. Multiple giant cell lesions in patients with Noonan syndrome and cardio-facio-cutaneous syndrome. Eur J Hum Genet 2009: 17: Happle R. Nevus sebaceus is a mosaic RASopathy. J Invest Dermatol 2013: 133: Campbell IM, Shaw CA, Stankiewicz P, Lupski JR. Somatic mosaicism: implications for disease and transmission genetics. Trends Genet 2015: 31: Hunter JC, Manandhar A, Carrasco MA, Gurbani D, Gondi S, Westover KD. Biochemical and structural analysis of common cancer-associated KRAS mutations. Mol Cancer Res 2015: 13: Vetter IR, Wittinghofer A. The guanine nucleotide-binding switch in three dimensions. Science 2001: 294: Feig LA, Cooper GM. Relationship among guanine nucleotide exchange, GTP hydrolysis, and transforming potential of mutated ras proteins. Mol Cell Biol 1988: 8: Akagi K, Uchibori R, Yamaguchi K, Kurosawa K, Tanaka Y, Kozu T. Characterization of a novel oncogenic K-ras mutation in colon cancer. Biochem Biophys Res Commun 2007: 352: Smith G, Bounds R, Wolf H, Steele RJ, Carey FA, Wolf CR. Activating K-Ras mutations outwith hotspot codons in sporadic colorectal tumours implications for personalised cancer medicine. Br J Cancer 2010: 102: Prior IA, Lewis PD, Mattos C. A comprehensive survey of Ras mutations in cancer. Cancer Res 2012: 72: Edkins S, O Meara S, Parker A et al. Recurrent KRAS codon 146 mutations in human colorectal cancer. Cancer Biol Ther 2006: 5: Kinsler VA, Krengel S, Riviere JB et al. Next-generation sequencing of nevus spilus-type congenital melanocytic nevus: exquisite genotype-phenotype correlation in mosaic RASopathies. J Invest Dermatol 2014: 134: Krengel S, Widmer DS, Kerl K, Levesque MP, Schiestl C, Weibel L. Naevus spilus-type congenital melanocytic naevus associated with a novel NRAS codon 61 mutation. Br J Dermatol 2016: 174: Groesser L, Herschberger E, Ruetten A et al. Postzygotic HRAS and KRAS mutations cause nevus sebaceous and Schimmelpenning syndrome. Nat Genet 2012: 44: Levinsohn JL, Tian LC, Boyden LM et al. Whole-exome sequencing reveals somatic mutations in HRAS and KRAS, which cause nevus sebaceus. J Invest Dermatol 2013: 133: Sun BK, Saggini A, Sarin KY et al. Mosaic activating RAS mutations in nevus sebaceus and nevus sebaceus syndrome. J Invest Dermatol 2013: 133:

9 Boppudi et al. 38. Hsu CK, Saito R, Nanda A et al. Systematised naevus sebaceus resulting from post-zygotic mutation in HRAS. Australas J Dermatol 2015, doi: /ajd [Epub ahead of print]. 39. Kim SW, Song JS, Kang MS, Sin JB, Ki CS, Jeon GW. Identification of somatic KRAS mutation in a Korean baby with nevus sebaceus syndrome. Ann Lab Med 2015: 35: Kuroda Y, Ohashi I, Enomoto Y et al. A postzygotic NRAS mutation in a patient with Schimmelpenning syndrome. Am J Med Genet A 2015: 167A: Wang H, Qian Y, Wu B, Zhang P, Zhou W. KRAS G12D mosaic mutation in a Chinese linear nevus sebaceous syndrome infant. BMC Med Genet 2015: 16: Hafner C, Toll A, Gantner S et al. Keratinocytic epidermal nevi are associated with mosaic RAS mutations. J Med Genet 2012: 49: Watanabe Y, Shido K, Niihori T et al. Somatic BRAF c.1799t>a p.v600e Mosaicism syndrome characterized by a linear syringocystadenoma papilliferum, anaplastic astrocytoma, and ocular abnormalities. Am J Med Genet A 2016: 170: Groesser L, Herschberger E, Sagrera A et al. Phacomatosis pigmentokeratotica is caused by a postzygotic HRAS mutation in a multipotent progenitor cell. J Invest Dermatol 2013: 133: Sarin KY, McNiff JM, Kwok S, Kim J, Khavari PA. Activating HRAS mutation in nevus spilus. J Invest Dermatol 2014: 134: Charbel C, Fontaine RH, Malouf GG et al. NRAS mutation is the sole recurrent somatic mutation in large congenital melanocytic nevi. J Invest Dermatol 2014: 134: Kinsler VA, Thomas AC, Ishida M et al. Multiple congenital melanocytic nevi and neurocutaneous melanosis are caused by postzygotic mutations in codon 61 of NRAS. J Invest Dermatol 2013: 133: Salgado CM, Basu D, Nikiforova M et al. BRAF mutations are also associated with neurocutaneous melanocytosis and large/giant congenital melanocytic nevi. Pediatr Dev Pathol 2015: 18: Levinsohn JL, Teng J, Craiglow BG et al. Somatic HRAS p.g12s mutation causes woolly hair and epidermal nevi. J Invest Dermatol 2014: 134: Lim YH, Ovejero D, Sugarman JS et al. Multilineage somatic activating mutations in HRAS and NRAS cause mosaic cutaneous and skeletal lesions, elevated FGF23 and hypophosphatemia. Hum Mol Genet 2014: 23: