ORIGINAL STUDY
Large Germline Deletions of the CYLD Gene in Patients With Brooke–Spiegler Syndrome and Multiple Familial Trichoepithelioma Tomas Vanecek, PhD,* Zbynek Halbhuber, PhD,† Denisa Kacerovska, MD, PhD,*‡ Petr Martinek, MSc,* Monika Sedivcova, MSc,* Richard A. Carr, FRCP,§ David Slouka, MD,¶ Michal Michal, MD,‡ and Dmitry V. Kazakov, MD, PhD‡
Abstract: Brooke–Spiegler syndrome (BSS) and its phenotypic variants, multiple familial trichoepithelioma (MFT) and familial cylindromatosis, are rare autosomal dominant hereditary diseases. They are characterized by the presence of multiple adnexal tumors, especially cylindromas, spiradenomas, spiradenocylindromas, and trichoepitheliomas. Implicated in the pathogenesis of the disease is the gene CYLD, which is localized on the long arm of chromosome 16. This gene encodes an evolutionarily conserved protein belonging to the deubiquitinating enzymes family, which plays a key role in many signaling pathways, especially in NF-kB, JNK, and Wnt. Less than 90 germline mutations of CYLD have been identified in patients with BSS/MFT. These mutations are mostly small alterations in the coding sequence and at exon–intron junction sites. One patient with an intronic mutation and another with a large CYLD deletion have also been recorded. In this study, the authors have analyzed a cohort of 14 patients with BSS/MFT from 13 families for large genome rearrangements by array comparative genome hybridization followed by confirmatory sequencing. We identified 2 large deletions, namely c.-34111_*297858del378779 and c.914-6398_1769del13642ins20 in patients with MFT and BSS, respectively. All other analyzable patients did not reveal any copy number alteration. It is concluded that the large rearrangements are relatively rare in patients without a germline CYLD mutation demonstrable by conventional sequencing. The pathogenetic mechanisms in patients with BSS/MFT lacking germline sequence alterations or large rearrangements in the CYLD gene remain to be clarified. Key Words: Brooke–Spiegler syndrome, multiple familial trichoepithelioma, trichoblastoma, spiradenoma, spiradenocylindroma, cylindroma, cylindromatosis, CYLD (Am J Dermatopathol 2014;36:868–874)
From the *Bioptical Laboratory, Pilsen, Czech Republic; †Central European Biosystems, Prague, Czech Republic; ‡Sikl’s Department of Pathology, Medical Faculty in Pilsen, Charles University in Prague, Pilsen, Czech Republic; §Department of Pathology, Warwick Hospital, Warwick, United Kingdom; and ¶Ear, Nose, and Throat Clinic, Medical Faculty in Pilsen, Charles University in Prague, Pilsen, Czech Republic. The authors declare no conflicts of interest. Reprints: Dmitry V. Kazakov, MD, PhD, Sikl’s Department of Pathology, Charles University Medical Faculty Hospital, Alej Svobody 80, 304 60 Pilsen, Czech Republic (e-mail: [email protected]). © 2014 Lippincott Williams & Wilkins
868
| www.amjdermatopathology.com
INTRODUCTION Brooke–Spiegler syndrome (BSS) and its phenotypic variants, multiple familial trichoepithelioma (MFT) and familial cylindromatosis, are rare autosomal dominant hereditary diseases characterized by multiple benign skin adnexal tumors, including cylindromas, spiradenomas, spiradenocylindromas, and trichoepitheliomas.1–14 These cutaneous adnexal neoplasms may infrequently undergo malignant transformation.15–19 Morphologically similar neoplasms may rarely arise in the salivary glands20–23 and exceptionally in the breast.24,25 Alterations of the CYLD gene are detected in a majority of familial cases of BSS and allelic disorders.26–29 This tumor suppressor gene is localized on the long arm of chromosome 16, has 20 exons, and encodes a 120 kDa evolutionarily conserved protein. This protein belongs to the deubiquitinating enzymes family and contains the following domains: cytoskeleton-associated glycine-rich, NEMO binding, phosphorylation, and ubiquitin (Ub)-specific protease (USP). CYLD plays a key role, as an inhibitor protein, in the NF-kB, JNK, and Wnt signaling pathways.3,26,30–35 Various germline mutations in CYLD have been found in BSS/MFT, mostly involving the coding sequence and exon–intron junction sites, especially in the regions where conserved enzymatic domains are encoded.3,36–40 Rarely, a germline CYLD mutation in a deep intronic region41 or a large germline CYLD deletion have been reported.42 Recently, we have published a cohort of 67 patients from 48 families affected by BSS/MFT, in which we studied CYLD germline and somatic mutations in coding sequence and exon–intron junction of the gene.36 In that study, we found the germline CYLD mutations in 76% and 44% of patients with BSB and MFT, respectively. This detection rate is in line with a previous smaller study using a similar methodology.43 Several possibilities have been suggested to explain the absence of a detectable CYLD mutation, including the presence of a large rearrangement (deletion, duplication, and others), methylation of the promoter, incorrect regulation on the expression level, and although this is the least likely a mutation in another gene. We undertook this study to confirm or exclude the possibility of large genome rearrangements in the CYLD gene region in patients with BSS/MFT, without a detectable sequence alteration of the gene, using high-resolution array comparative genome hybridization (aCGH) at chromosome 16. Am J Dermatopathol Volume 36, Number 11, November 2014
Am J Dermatopathol Volume 36, Number 11, November 2014
MATERIALS AND METHODS
Brooke–Spiegler Syndrome
Array Comparative Genome Hybridization Microarray Processing
Case Selection In the previously published cohort of 67 patients affected by BSS/MFT, there were 16 patients from 12 families with no germline CYLD gene mutation detected by the sequencing of coding exons and exon–intron junctions.36 One new patient with BSS negative for CYLD mutation by conventional sequencing was added. The 17 patients, belonging to 13 families, included 7 manifesting the classic BSS phenotype and 10 with MFT (Table 1). One case per family was tested, with the exception of family 7 (cases 10 and 11), in which 2 family members were investigated, that is, 14 individuals from 13 families (Table 1) (Figs. 1, 2).
DNA Extraction DNA from the peripheral blood (10 cases) and nontumor formalin-fixed paraffin-embedded tissue (4 cases) (Table 1) was extracted using the NucleoSpin Tissue Kit (Macherey Nagel, Duren, Germany) according to the manufacturers’ protocol.
DNA Quantity, Purity, and Integrity The quantity and purity of the isolated nucleic acids were assessed with a Nanodrop 1000 spectrophotometer (Thermo Scientific, Waltham, MA). DNA integrity was determined by a multiplex polymerase chain reaction (PCR) amplifying 5 fragments ranging from 100 to 600 bp followed by analysis on 2% agarose (Serva, Heidelberg, Germany) gel electrophoresis.44 Samples with A260/280 nm absorbance ratio between 1.8 and 2 and showing integrity of DNA allowing amplification of all the PCR fragments were used for aCGH.
A Nimblegen (Roche NimbleGen Inc., Madison, WI) Custom CGH 6 · 630 K Array chromosome 16 was used for analysis.45 A custom design of hg19 chromosome 16 was assembled using a median probe spacing of 1 probe starting every 100 bases with respect to the maximal coverage of the CYLD gene, where only Short Interspersed Elements and Long Interspersed Elements repetitive elements were omitted. To provide better coverage in the telomeric and centromeric regions of the chromosome, the selected probes were allowed to match up to 10 times in the genome. Five hundred nanogram of gDNA was labeled with a NimbleGen standard protocol (NimbleGen Arrays User’s Guide version 9.1), which includes Cy3 labeling of the test sample and Cy5 labeling of the reference sample (MegaPool Reference DNA Male, MegaPool Reference DNA Female; Kreatech Diagnostics, Amsterdam, the Netherlands). The reference and the test sample were mixed, dried, and hybridized for 60 hours at 428C in the mix mode B using MAUI Hybridization system (Roche NimbleGen Inc). Posthybridization washing was done using NimbleGen buffers with increasing stringency (Roche NimbleGen Inc) followed by the microarray scanning with InnoScan 900 (Innopsys, Carbonne, France) at resolution 1 mm. Image analysis was performed using the NimbleScan 2.6 software (Roche NimbleGen Inc) according to appropriate .ndf file.
Data Analysis The data analysis was also processed in a NimbleScan 2.6 (Roche NimbleGen Inc) using CGH-segMNT analysis. Before segmentation analysis, the qspline fit normalization that
TABLE 1. The Main Clinicopathological Features and Results of Molecular Biologic Studies in 17 BSS/MFT Patients With Lack of a Demonstrable CYLD Germline Mutation Family Number 1 2 3 4 5 6 7
8 9 10 11 12 13
Patient Sample Sex/ Nationality/Ethnic Source of aCGH Number Phenotype Age Background DNA Analysis 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
MFT MFT MFT BSS BSS BSS MFT MFT MFT MFT MFT BSS BSS MFT MFT BSS BSS
M/36 F/30 M/46 F/59 M/38 F/61 F/28 M/21 F/14 M/38 F/19 F/43 F/50 F/49 F/20 F/47 M/64
Russian Czech German Czech Czech Spanish Spanish USA Czech Czech Czech Czech Belorussian Austrian German Czech English
B B FFPE B B FFPE B FFPE B B B FFPE B B B B FFPE
P P NA P NP NP P P NP P P NA P P NA P P
aCGH Result
Mutation Description
Deletion c.-34111_*297858del378779 Negative — — — Negative — — — — — Negative — Negative — — — Negative — Negative — — — Negative — Negative — — — Deletion c.914-6398_1769del13642ins20 Negative —
Number of Family/Patient in Grossmann et al, 2013 37/52 38/53 39/54 40/55 40/56 41/57 41/58 42/59 43/60 43/61 43/62 44/63 45/64 46/65 47/66 48/67 New case
B, blood; FFPE, formalin-fixed paraffin-embedded tissue; NP, not performed; NA, not analyzable; P, performed.
2014 Lippincott Williams & Wilkins
www.amjdermatopathology.com |
869
Vanecek et al
Am J Dermatopathol Volume 36, Number 11, November 2014
exported into .gpr, .txt, and .pdf files to allow further viewing and analysis of data in SignalMap (Roche NimbleGen Inc), general text editor or tab processor. Data were annotated in respect to the reference genome hg19 (Genome Reference Consortium Human Build 37).
Rearrangement Spanning PCR For exact determination of rearrangements, primers surrounding the affected region were designed for each suspect case (see Results and Table 2). All primers were designed using the Primer 3 software (http://frodo.wi.mit. edu/cgi-bin/primer3/primer3_www.cgi)46 and human chromosome 16 sequence (version GRCh37/hg19). Then, 100 ng of DNA was added to a mixture consisting of 10 pmole of forward and reverse primers (Table 2), 12.5 mL HotStar Taq PCR Master Mix (QIAgen, Hilden, Germany), and distilled water up to 25 mL. Initial denaturation at 958C for 14 minutes was followed by 40 cycles of denaturation at 958C for 1 minute, annealing at 608C for 1 minute, and extension at 728C for 2.5 minutes. The program was terminated by incubation at 728C for 10 minutes. The reactions were analyzed with agarose (Serva) gel electrophoresis. Successfully amplified PCR products were purified with magnetic beads using Agencourt AMPure kit (Agencourt Bioscience Corporation, A Beckman Coulter Company, Beverly, MA), both side sequenced using Big Dye Terminator Sequencing kit (Applied Biosystems, Carlsbad, CA), and purified with magnetic beads using Agencourt CleanSEQ kit (Agencourt Bioscience Corporation), all according to the manufacturer’s instructions. Samples were then run on an automated sequencer ABI Prism 3130xl (Applied Biosystems) at a constant voltage of 13.2 kV for 20 minutes. Data were processed in Sequencing Analysis software 5.3.1 (Applied Biosystems), and received sequences were compared with chromosome 16 reference sequence using BLAST software (http://blast.ncbi.nlm.nih.gov/Blast.cgi). Nucleotides from the targeted region were numbered using CYLD coding DNA reference sequence (NG_012061.1), and the ascertained mutations were described using current mutation nomenclature (http://www.hgvs.org/mutnomen/recs.html). FIGURE 1. Case 1. Clinicopathological features of the patient with MFTs. Multiple cutaneous nodules on the face (A) histopathologically representing trichoepitheliomas [cribriform trichoblastoma (B, C)].
compensates for inherent differences in signal between the 2 dyes was applied. The segMNT algorithm identifies copy number changes using a dynamic programming process that minimizes the squared error relative to the segment means. This procedure allows one to generate a list of candidate breakpoints, to identify the best segmentation for each given number of and to determine the number of segments to output results. The minimum segment difference in the log2 ratio that 2 segments must exhibit before they are identified as separate segments was set to 0.1. The minimum segment length was set to 2 probes. The stringency with which the initial segment boundaries were selected was set to the maximal stringent value 0.9. The segmentation nonaveraged and reduced averaging window was applied to raw data, which gave us data spacing of 100 and 100,000 bp, respectively. The results were
870
| www.amjdermatopathology.com
RESULTS First, we analyzed quantity, purity, and integrity of the DNA extracted from the patient’s peripheral blood and nontumor tissue samples using spectrophotometry and PCR, respectively. Eleven of the 14 specimens (9 extracted from blood, 2 extracted from formalin-fixed paraffin-embedded tissue) met the set criteria (Table 1). These 11 specimens were analyzed by high-resolution aCGH of chromosome 16. The aCGH analysis revealed a higher signal in the reference channel (red) comparing with a test channel (green) in a CYLD gene region, indicating a large deletion in 2 cases, namely cases 1 and 16 (Fig. 3). In case 1, a decrease in signal intensity in green channel started at probe position 50, 749, 626 and ended at probe position 51, 128, 051 on chromosome 16. In case 16, a decrease in green channel intensity started at probe position 50, 802, 675 and ended at probe position 50, 816, 050 on chromosome 16. Several sets of primers with increasing 2014 Lippincott Williams & Wilkins
Am J Dermatopathol Volume 36, Number 11, November 2014
Brooke–Spiegler Syndrome
TABLE 2. Primers Used for Rearrangement Spanning PCR Case
Name
Case 1 primer pairs
F1 R1 F2 R2 F3 R3 F3 R5 F6 R3 F6 R5 F8 R3 F8 R5 F10 R10 F1 R1 F3 R3 F4 R4 F4 R5 F6 R6 F6 R7 F8 R8 F2 R7 F8 R7
Case 16 primer pairs
Sequence 50 -30 GTCACTTCGCTTGTCTGTGC GCCTTCCATTGTGCATGATT GGCCATAAGAGGCCCTGAG AAGATGAAATGCCTCAAATTCC TGTTTCCATGCCAACTGAAA CTCAGGGAACCGTGAATGTT TGTTTCCATGCCAACTGAAA AAGAAGGGGGTTTTCCTCAA CAACCAGCTTGGTTTTCCAT CTCAGGGAACCGTGAATGTT CAACCAGCTTGGTTTTCCAT AAGAAGGGGGTTTTCCTCAA TCCTCATAGGCCAGTCCATC CTCAGGGAACCGTGAATGTT TCCTCATAGGCCAGTCCATC AAGAAGGGGGTTTTCCTCAA TGAGTGGGAATTTGGAAACC CCCCCAGGTTTGTGTCTTTA TGGCTTCTGAAGGATACCTTTG GCCTCCAAATGCTGAAAAGA CATCTGCCTTTCATGTCATCAT GGTATGTGTCCCTGCCTCAT TGGCTCAGAGCAGTGAAGAG CACAGAAACATGAGACAACATTCA TGGCTCAGAGCAGTGAAGAG TGCATATATGAGGGCAGCAG CGTCACTTGTCAGGCACACT GGCCGCCACTTTATTCTTCT CGTCACTTGTCAGGCACACT CTATTCAGTGCCCCACTTCC TTGAAGTGCTTTCTCCCACA AATTAGCCAGGCATGATGGA TTGCCCAATTACCTTCATCC CTATTCAGTGCCCCACTTCC TTGAAGTGCTTTCTCCCACA CTATTCAGTGCCCCACTTCC
FIGURE 2. Case 16. Clinicopathological features of the patient with BSS. Multiple cutaneous nodules on the forehead (A) histopathologically representing spiradenocylindromas (B, C).
on agarose gel electrophoresis. The lengths of products were of an approximately expected size (Fig. 4). All PCR fragments were then both side sequenced, and the exact positions of the rearrangements were defined. In case 1, a mutation c.34111_*297858del378779 (g.50749499_51128277del378779) was revealed. This deletion starts in intron 4 of the adjacent gene NOD2 and ends approximately 300 kB downstream of the stop codon in the CYLD gene.47 In case 16, a mutation c.914-6398_1769del13642ins20 (g.50802679_50816320del13642ins20) was found. The latter deletion starts in intron 6, ends in exon 12, of the CYLD gene and is accompanied by a small insertion of 20 bp of unknown origin (searched by BLAST) (Fig. 5; Table 1).
amplification range surrounding each rearrangement position were designed for each case (Table 2), and both DNA samples were amplified. After PCR amplification, 9 of 9 fragments in case 1 and 8 of 9 fragments in case 16 were detected
We analyzed a cohort of patients with BSS/MFT with the previously nondetectable germline sequence mutation in
2014 Lippincott Williams & Wilkins
DISCUSSION
www.amjdermatopathology.com |
871
Vanecek et al
Am J Dermatopathol Volume 36, Number 11, November 2014
FIGURE 3. The results of aCGH. A, The position of NOD2, CYLD, and SALL1 on chromosme 16. B, An approximately 0.4 MB deletion in case 1. C, An approximately 13.6 kB deletion in case 16.
coding region of the CYLD gene36 to search for large rearrangements, using aCGH followed by sequencing of rearrangement spanning PCR products in positive samples. Although a somatic mutation such as loss of heterozygosity caused by, among other things, a large deletion is apparently a common feature in patients with BSS/MFT,26,48,49,36 a germline alteration represented by a large deletion seems to be rare. To date, only 1 patient with BSS (reported as having familial cylindromatosis) with a large germline rearrangement in the CYLD gene has been described.42 This was an approximately 5.3-kB deletion involving and extending beyond the 30 end of the gene, affecting the USP domain of the CYLD protein. This patient was identified among 13 cases of BSS included in that study.42,3 In our study group, we found 2 novel large germline deletions in 2 patients, 1 with classical BSS phenotype and 1 with
MFT, representing 18% (2 of 14) of CYLD sequence mutation– negative analyzable cases and 3% (2 of 68) of the cohort. One of the 2 deletions was accompanied by a tiny insertion, thus technically representing a delins mutation. When the clinical phenotype is taken into account, large germline deletions are found in 2% and 5% of patients with classic BSS and MFT, respectively. No other large rearrangements (eg, duplication, insertion, and unbalanced translocation) were found in our cohort. The frequency of large rearrangements seems to be slightly lower when compared with other tumor suppressor genes affected in syndromes manifested by skin tumors, including Lynch syndrome/Muir–Torre syndrome and neurofibromatosis type 1 but comparable with tuberous sclerosis, Gorlin–Goltz, and Cowden syndromes.50–56 In our cohort, there is still a relatively high percentage of patients without detectable germline
FIGURE 4. The results of rearrangement spanning PCR. A, PCR products of expected sizes detected in 9 of 9 reactions in case 1. B, PCR products of expected sizes detected in 8 of 9 reactions in case 16. Normal control DNA amplified with designed primer sets and nontemplate controls were included but not shown. All control reaction were negative.
872
| www.amjdermatopathology.com
2014 Lippincott Williams & Wilkins
Am J Dermatopathol Volume 36, Number 11, November 2014
Brooke–Spiegler Syndrome
FIGURE 5. The results of sequencing analysis of selected PCR products. A, Arrow shows the position of the mutation in case 1. B, Arrows show the position of the mutation in case 16. Twenty base pair insert is also depicted.
mutation in the CYLD gene. Mutations in the coding region36 and large rearrangements have been excluded leaving alterations in intronic41 and regulatory regions, balanced translocations and, perhaps, epigenetic silencing, as possible explanations for BSS/MFT in our cohort. Less likely considerations for the pathogenesis of some BSS/MFT cases include modification at the expression level (eg, through microRNA) or another gene or regulatory pathway. The large molecular alterations found in 2 of our cases are quite different. The deletion (or technically delins mutation) in case 16 alters only approximately 13.5 kB of the CYLD gene, including the region encoding the third cytoskeleton-associated glycine-rich domain and the beginning of the USP domain. The mutation in case 1 is much more extensive, amounting to near complete deletion of the CYLD gene and the surrounding regions (including regulatory), affecting also a part of NOD2. The NOD2 gene is a member of the Nod1/Apaf-1 family and encodes a protein with 2 caspase recruitment domains and 6 leucine-rich repeats.57 NOD2 plays a role in the immune response, activating NF-kB signaling pathway, after recognition of intracellular bacterial lipopolysaccharides. Mutations in NOD2 have been associated with Crohn disease and Blau syndrome.58 No information was available to confirm or exclude the latter possibilities in our patient. In conclusion, we have identified 2 novel large germline rearrangement mutations including 1 pure deletion and 1 deletion/insertion (delins mutation) in a cohort of BSS/MFT. We have also demonstrated that large rearrangements are relatively rare in these patients. Other pathogenetic mechanisms in BSS/MFT, lacking germline sequence alterations, or large rearrangements in the CYLD gene, remain to be clarified. REFERENCES 1. Biggs PJ, Wooster R, Ford D, et al. Familial cylindromatosis (turban tumour syndrome) gene localised to chromosome 16q12-q13: evidence for its role as a tumour suppressor gene. Nat Genet. 1995;11:441–443.
2014 Lippincott Williams & Wilkins
2. Kazakov DV, Schaller J, Vanecek T, et al. Brooke-Spiegler syndrome: report of a case with a novel mutation in the CYLD gene and different types of somatic mutations in benign and malignant tumors. J Cutan Pathol. 2010;37:886–890. 3. Blake PW, Toro JR. Update of cylindromatosis gene (CYLD) mutations in Brooke-Spiegler syndrome: novel insights into the role of deubiquitination in cell signaling. Hum Mutat. 2009;30:1025–1036. 4. Kazakov DV, Soukup R, Mukensnabl P, et al. Brooke-Spiegler syndrome: report of a case with combined lesions containing cylindromatous, spiradenomatous, trichoblastomatous, and sebaceous differentiation. Am J Dermatopathol. 2005;27:27–33. 5. Bowen S, Gill M, Lee DA, et al. Mutations in the CYLD gene in BrookeSpiegler syndrome, familial cylindromatosis, and multiple familial trichoepithelioma: lack of genotype-phenotype correlation. J Invest Dermatol. 2005;124:919–920. 6. Ponti G, Nasti S, Losi L, et al. Brooke-Spiegler syndrome: report of two cases not associated with a mutation in the CYLD and PTCH tumorsuppressor genes. J Cutan Pathol. 2012;39:366–371. 7. Kazakov DV, Michal M, Kacerovska D, et al. Cutaneous adnexal tumors. Philadelphia, PA: Lippincott Williams & Wilkins; p.814. 2012. 8. Weyers W, Nilles M, Eckert F, et al. Spiradenomas in Brooke-Spiegler syndrome. Am J Dermatopathol. 1993;15:156–161. 9. Kazakov DV, Vanecek T, Zelger B, et al. Multiple (familial) trichoepitheliomas: a clinicopathological and molecular biological study, including CYLD and PTCH gene analysis, of a series of 16 patients. Am J Dermatopathol. 2011;33:251–265. 10. Clarke J, Ioffreda M, Helm KF. Multiple familial trichoepitheliomas: a folliculosebaceous-apocrine genodermatosis. Am J Dermatopathol. 2002;24:402–405. 11. Puig L, Nadal C, Fernandez-Figueras MT, et al. Brooke-Spiegler syndrome variant: segregation of tumor types with mixed differentiation in two generations. Am J Dermatopathol. 1998;20:56–60. 12. Petersson F, Kutzner H, Spagnolo DV, et al. Adenoid cystic carcinomalike pattern in spiradenoma and spiradenocylindroma: a rare feature in sporadic neoplasms and those associated with Brooke-Spiegler syndrome. Am J Dermatopathol. 2009;31:642–648. 13. Kazakov DV, Vanecek T, Nemcova J, et al. Spectrum of tumors with follicular differentiation in a patient with the clinical phenotype of multiple familial trichoepitheliomas: a clinicopathological and molecular biological study, including analysis of the CYLD and PTCH genes. Am J Dermatopathol. 2009;31:819–827. 14. Uede K, Yamamoto Y, Furukawa F. Brooke-Spiegler syndrome associated with cylindroma, trichoepithelioma, spiradenoma, and syringoma. J Dermatol. 2004;31:32–38. 15. Pizinger K, Michal M. Malignant cylindroma in Brooke-Spiegler syndrome. Dermatology. 2000;201:255–257.
www.amjdermatopathology.com |
873
Vanecek et al
16. Kazakov DV, Zelger B, Rutten A, et al. Morphologic diversity of malignant neoplasms arising in preexisting spiradenoma, cylindroma, and spiradenocylindroma based on the study of 24 cases, sporadic or occurring in the setting of Brooke-Spiegler syndrome. Am J Surg Pathol. 2009;33: 705–719. 17. De Francesco V, Frattasio A, Pillon B, et al. Carcinosarcoma arising in a patient with multiple cylindromas. Am J Dermatopathol. 2005;27:21–26. 18. Antonescu CR, Terzakis JA. Multiple malignant cylindromas of skin in association with basal cell adenocarcinoma with adenoid cystic features of minor salivary gland. J Cutan Pathol. 1997;24:449–453. 19. Pincus LB, McCalmont TH, Neuhaus IM, et al. Basal cell carcinomas arising within multiple trichoepitheliomas. J Cutan Pathol. 2008;35 (suppl 1):59–64. 20. Pingitore R, Campani D. Salivary gland involvement in a case of dermal eccrine cylindroma of the scalp (turban tumor). Report of a case with lung metastases. Tumori. 1984;70:385–388. 21. Kakagia D, Alexiadis G, Kiziridou A, et al. Brooke-Spiegler syndrome with parotid gland involvement. Eur J Dermatol. 2004;14:139–141. 22. Kostler E, Schonlebe J, Mentzel T, et al. Psoriasis and Brooke-Spiegler syndrome with multiple malignancies. J Eur Acad Dermatol Venereol. 2005;19:380–381. 23. Scott AR, Faquin WC, Deschler DG. Parotid mass in a woman with multiple cutaneous cylindromas. Head Neck. 2010;32:684–687. 24. Nonaka D, Rosai J, Spagnolo D, et al. Cylindroma of the breast of skin adnexal type: a study of 4 cases. Am J Surg Pathol. 2004;28:1070–1075. 25. Kazakov DV, Spagnolo DV, Kacerovska D, et al. Cutaneous type adnexal tumors outside the skin. Am J Dermatopathol. 2011;33:303–315. 26. Bignell GR, Warren W, Seal S, et al. Identification of the familial cylindromatosis tumour-suppressor gene. Nat Genet. 2000;25:160–165. 27. Thomson SA, Rasmussen SA, Zhang J, et al. A new hereditary cylindromatosis family associated with CYLD1 on chromosome 16. Hum Genet. 1999;105:171–173. 28. Lv HL, Huang YJ, Zhou D, et al. A novel missense mutation of CYLD gene in a Chinese family with multiple familial trichoepithelioma. J Dermatol Sci. 2008;50:143–146. 29. Oiso N, Mizuno N, Fukai K, et al. Mild phenotype of familial cylindromatosis associated with an R758X nonsense mutation in the CYLD tumour suppressor gene. Br J Dermatol. 2004;151:1084–1086. 30. Biggs PJ, Chapman P, Lakhani SR, et al. The cylindromatosis gene (cyld1) on chromosome 16q may be the only tumour suppressor gene involved in the development of cylindromas. Oncogene. 1996;12: 1375–1377. 31. Massoumi R, Paus R. Cylindromatosis and the CYLD gene: new lessons on the molecular principles of epithelial growth control. Bioessays. 2007; 29:1203–1214. 32. Masoumi KC, Shaw-Hallgren G, Massoumi R. Tumor suppressor function of CYLD in nonmelanoma skin cancer. J Skin Cancer. 2011;2011: 614097. 33. Tauriello DV, Haegebarth A, Kuper I, et al. Loss of the tumor suppressor CYLD enhances Wnt/beta-catenin signaling through K63-linked ubiquitination of Dvl. Mol Cell. 2010;37:607–619. 34. Young AL, Kellermayer R, Szigeti R, et al. CYLD mutations underlie Brooke-Spiegler, familial cylindromatosis, and multiple familial trichoepithelioma syndromes. Clin Genet. 2006;70:246–249. 35. Wickstrom SA, Masoumi KC, Khochbin S, et al. CYLD negatively regulates cell-cycle progression by inactivating HDAC6 and increasing the levels of acetylated tubulin. EMBO J. 2010;29:131–144. 36. Grossmann P, Vanecek T, Steiner P, et al. Novel and recurrent germline and somatic mutations in a cohort of 67 patients from 48 families with Brooke-Spiegler syndrome including the phenotypic variant of multiple familial trichoepitheliomas and correlation with the histopathologic findings in 379 biopsy specimens. Am J Dermatopathol. 2013;35:34–44. 37. Almeida S, Maillard C, Itin P, et al. Five new CYLD mutations in skin appendage tumors and evidence that aspartic acid 681 in CYLD is essential for deubiquitinase activity. J Invest Dermatol. 2008;128:587–593.
874
| www.amjdermatopathology.com
Am J Dermatopathol Volume 36, Number 11, November 2014
38. Kacerovska D, Szep Z, Kollarikova L, et al. A novel germline mutation in the CYLD gene in a Slovak patient with Brooke-Spiegler syndrome. Cesk Patol. 2013;49:89–92. 39. Zheng G, Hu L, Huang W, et al. CYLD mutation causes multiple familial trichoepithelioma in three Chinese families. Hum Mutat. 2004;23:400. 40. Linos K, Schwartz J, Kazakov DV, et al. Recurrent CYLD nonsense mutation associated with a severe, disfiguring phenotype in an African American family with multiple familial trichoepithelioma. Am J Dermatopathol. 2011;33:640–642. 41. Kazakov DV, Thoma-Uszynski S, Vanecek T, et al. A case of BrookeSpiegler syndrome with a novel germline deep intronic mutation in the CYLD gene leading to intronic exonization, diverse somatic mutations, and unusual histology. Am J Dermatopathol. 2009;31:664–673. 42. van den Ouweland AM, Elfferich P, Lamping R, et al. Identification of a large rearrangement in CYLD as a cause of familial cylindromatosis. Fam Cancer. 2010;10:127–132. 43. Saggar S, Chernoff KA, Lodha S, et al. CYLD mutations in familial skin appendage tumours. J Med Genet. 2008;45:298–302. 44. van Dongen JJ, Langerak AW, Bruggemann M, et al. Design and standardization of PCR primers and protocols for detection of clonal immunoglobulin and T-cell receptor gene recombinations in suspect lymphoproliferations: report of the BIOMED-2 Concerted Action BMH4-CT98-3936. Leukemia. 2003;17:2257–2317. 45. Workman C, Jensen LJ, Jarmer H, et al. A new non-linear normalization method for reducing variability in DNA microarray experiments. Genome Biol. 2002;3:research0048. 46. Rozen S, Skaletsky H. Primer3 on the WWW for general users and for biologist programmers. Methods Mol Biol. 2000;132:365–386. 47. Moreira LO, Zamboni DS. NOD1 and NOD2 signaling in infection and inflammation. Front Immunol. 2012;3:328. 48. Leonard N, Chaggar R, Jones C, et al. Loss of heterozygosity at cylindromatosis gene locus, CYLD, in sporadic skin adnexal tumours. J Clin Pathol. 2001;54:689–692. 49. Sima R, Vanecek T, Kacerovska D, et al. Brooke-Spiegler syndrome: report of 10 patients from 8 families with novel germline mutations: evidence of diverse somatic mutations in the same patient regardless of tumor type. Diagn Mol Pathol. 2010;19:83–91. 50. Peltomaki P, Vasen H. Mutations associated with HNPCC predisposition— update of ICG-HNPCC/INSiGHT mutation database. Dis Markers. 2004; 20:269–276. 51. Strizheva GD, Carsillo T, Kruger WD, et al. The spectrum of mutations in TSC1 and TSC2 in women with tuberous sclerosis and lymphangiomyomatosis. Am J Respir Crit Care Med. 2001;163:253–258. 52. Brook-Carter PT, Peral B, Ward CJ, et al. Deletion of the TSC2 and PKD1 genes associated with severe infantile polycystic kidney disease— a contiguous gene syndrome. Nat Genet. 1994;8:328–332. 53. Kacerovska D, Vrtel R, Michal M, et al. TSC2/PKD1 contiguous gene syndrome: a report of 2 cases with emphasis on dermatopathologic findings. Am J Dermatopathol. 2009;31:532–541. 54. Longa L, Scolari F, Brusco A, et al. A large TSC2 and PKD1 gene deletion is associated with renal and extrarenal signs of autosomal dominant polycystic kidney disease. Nephrol Dial Transplant. 1997; 12:1900–1907. 55. Eng C. The role of PTEN, a phosphatase gene, in inherited and sporadic nonmedullary thyroid tumors. Recent Prog Horm Res. 1999;54:441–452; discussion 453. 56. Marsh DJ, Coulon V, Lunetta KL, et al. Mutation spectrum and genotype-phenotype analyses in Cowden disease and BannayanZonana syndrome, two hamartoma syndromes with germline PTEN mutation. Hum Mol Genet. 1998;7:507–515. 57. Hugot JP, Chamaillard M, Zouali H, et al. Association of NOD2 leucinerich repeat variants with susceptibility to Crohn’s disease. Nature. 2001; 411:599–603. 58. Abraham C, Cho JH. Functional consequences of NOD2 (CARD15) mutations. Inflamm Bowel Dis. 2006;12:641–650.
2014 Lippincott Williams & Wilkins
Large Germline Deletions of the CYLD Gene in Patients With Brooke–Spiegler Syndrome and Multiple Familial Trichoepithelioma Tomas Vanecek, PhD,* Zbynek Halbhuber, PhD,† Denisa Kacerovska, MD, PhD,*‡ Petr Martinek, MSc,* Monika Sedivcova, MSc,* Richard A. Carr, FRCP,§ David Slouka, MD,¶ Michal Michal, MD,‡ and Dmitry V. Kazakov, MD, PhD‡
Abstract: Brooke–Spiegler syndrome (BSS) and its phenotypic variants, multiple familial trichoepithelioma (MFT) and familial cylindromatosis, are rare autosomal dominant hereditary diseases. They are characterized by the presence of multiple adnexal tumors, especially cylindromas, spiradenomas, spiradenocylindromas, and trichoepitheliomas. Implicated in the pathogenesis of the disease is the gene CYLD, which is localized on the long arm of chromosome 16. This gene encodes an evolutionarily conserved protein belonging to the deubiquitinating enzymes family, which plays a key role in many signaling pathways, especially in NF-kB, JNK, and Wnt. Less than 90 germline mutations of CYLD have been identified in patients with BSS/MFT. These mutations are mostly small alterations in the coding sequence and at exon–intron junction sites. One patient with an intronic mutation and another with a large CYLD deletion have also been recorded. In this study, the authors have analyzed a cohort of 14 patients with BSS/MFT from 13 families for large genome rearrangements by array comparative genome hybridization followed by confirmatory sequencing. We identified 2 large deletions, namely c.-34111_*297858del378779 and c.914-6398_1769del13642ins20 in patients with MFT and BSS, respectively. All other analyzable patients did not reveal any copy number alteration. It is concluded that the large rearrangements are relatively rare in patients without a germline CYLD mutation demonstrable by conventional sequencing. The pathogenetic mechanisms in patients with BSS/MFT lacking germline sequence alterations or large rearrangements in the CYLD gene remain to be clarified. Key Words: Brooke–Spiegler syndrome, multiple familial trichoepithelioma, trichoblastoma, spiradenoma, spiradenocylindroma, cylindroma, cylindromatosis, CYLD (Am J Dermatopathol 2014;36:868–874)
From the *Bioptical Laboratory, Pilsen, Czech Republic; †Central European Biosystems, Prague, Czech Republic; ‡Sikl’s Department of Pathology, Medical Faculty in Pilsen, Charles University in Prague, Pilsen, Czech Republic; §Department of Pathology, Warwick Hospital, Warwick, United Kingdom; and ¶Ear, Nose, and Throat Clinic, Medical Faculty in Pilsen, Charles University in Prague, Pilsen, Czech Republic. The authors declare no conflicts of interest. Reprints: Dmitry V. Kazakov, MD, PhD, Sikl’s Department of Pathology, Charles University Medical Faculty Hospital, Alej Svobody 80, 304 60 Pilsen, Czech Republic (e-mail: [email protected]). © 2014 Lippincott Williams & Wilkins
868
| www.amjdermatopathology.com
INTRODUCTION Brooke–Spiegler syndrome (BSS) and its phenotypic variants, multiple familial trichoepithelioma (MFT) and familial cylindromatosis, are rare autosomal dominant hereditary diseases characterized by multiple benign skin adnexal tumors, including cylindromas, spiradenomas, spiradenocylindromas, and trichoepitheliomas.1–14 These cutaneous adnexal neoplasms may infrequently undergo malignant transformation.15–19 Morphologically similar neoplasms may rarely arise in the salivary glands20–23 and exceptionally in the breast.24,25 Alterations of the CYLD gene are detected in a majority of familial cases of BSS and allelic disorders.26–29 This tumor suppressor gene is localized on the long arm of chromosome 16, has 20 exons, and encodes a 120 kDa evolutionarily conserved protein. This protein belongs to the deubiquitinating enzymes family and contains the following domains: cytoskeleton-associated glycine-rich, NEMO binding, phosphorylation, and ubiquitin (Ub)-specific protease (USP). CYLD plays a key role, as an inhibitor protein, in the NF-kB, JNK, and Wnt signaling pathways.3,26,30–35 Various germline mutations in CYLD have been found in BSS/MFT, mostly involving the coding sequence and exon–intron junction sites, especially in the regions where conserved enzymatic domains are encoded.3,36–40 Rarely, a germline CYLD mutation in a deep intronic region41 or a large germline CYLD deletion have been reported.42 Recently, we have published a cohort of 67 patients from 48 families affected by BSS/MFT, in which we studied CYLD germline and somatic mutations in coding sequence and exon–intron junction of the gene.36 In that study, we found the germline CYLD mutations in 76% and 44% of patients with BSB and MFT, respectively. This detection rate is in line with a previous smaller study using a similar methodology.43 Several possibilities have been suggested to explain the absence of a detectable CYLD mutation, including the presence of a large rearrangement (deletion, duplication, and others), methylation of the promoter, incorrect regulation on the expression level, and although this is the least likely a mutation in another gene. We undertook this study to confirm or exclude the possibility of large genome rearrangements in the CYLD gene region in patients with BSS/MFT, without a detectable sequence alteration of the gene, using high-resolution array comparative genome hybridization (aCGH) at chromosome 16. Am J Dermatopathol Volume 36, Number 11, November 2014
Am J Dermatopathol Volume 36, Number 11, November 2014
MATERIALS AND METHODS
Brooke–Spiegler Syndrome
Array Comparative Genome Hybridization Microarray Processing
Case Selection In the previously published cohort of 67 patients affected by BSS/MFT, there were 16 patients from 12 families with no germline CYLD gene mutation detected by the sequencing of coding exons and exon–intron junctions.36 One new patient with BSS negative for CYLD mutation by conventional sequencing was added. The 17 patients, belonging to 13 families, included 7 manifesting the classic BSS phenotype and 10 with MFT (Table 1). One case per family was tested, with the exception of family 7 (cases 10 and 11), in which 2 family members were investigated, that is, 14 individuals from 13 families (Table 1) (Figs. 1, 2).
DNA Extraction DNA from the peripheral blood (10 cases) and nontumor formalin-fixed paraffin-embedded tissue (4 cases) (Table 1) was extracted using the NucleoSpin Tissue Kit (Macherey Nagel, Duren, Germany) according to the manufacturers’ protocol.
DNA Quantity, Purity, and Integrity The quantity and purity of the isolated nucleic acids were assessed with a Nanodrop 1000 spectrophotometer (Thermo Scientific, Waltham, MA). DNA integrity was determined by a multiplex polymerase chain reaction (PCR) amplifying 5 fragments ranging from 100 to 600 bp followed by analysis on 2% agarose (Serva, Heidelberg, Germany) gel electrophoresis.44 Samples with A260/280 nm absorbance ratio between 1.8 and 2 and showing integrity of DNA allowing amplification of all the PCR fragments were used for aCGH.
A Nimblegen (Roche NimbleGen Inc., Madison, WI) Custom CGH 6 · 630 K Array chromosome 16 was used for analysis.45 A custom design of hg19 chromosome 16 was assembled using a median probe spacing of 1 probe starting every 100 bases with respect to the maximal coverage of the CYLD gene, where only Short Interspersed Elements and Long Interspersed Elements repetitive elements were omitted. To provide better coverage in the telomeric and centromeric regions of the chromosome, the selected probes were allowed to match up to 10 times in the genome. Five hundred nanogram of gDNA was labeled with a NimbleGen standard protocol (NimbleGen Arrays User’s Guide version 9.1), which includes Cy3 labeling of the test sample and Cy5 labeling of the reference sample (MegaPool Reference DNA Male, MegaPool Reference DNA Female; Kreatech Diagnostics, Amsterdam, the Netherlands). The reference and the test sample were mixed, dried, and hybridized for 60 hours at 428C in the mix mode B using MAUI Hybridization system (Roche NimbleGen Inc). Posthybridization washing was done using NimbleGen buffers with increasing stringency (Roche NimbleGen Inc) followed by the microarray scanning with InnoScan 900 (Innopsys, Carbonne, France) at resolution 1 mm. Image analysis was performed using the NimbleScan 2.6 software (Roche NimbleGen Inc) according to appropriate .ndf file.
Data Analysis The data analysis was also processed in a NimbleScan 2.6 (Roche NimbleGen Inc) using CGH-segMNT analysis. Before segmentation analysis, the qspline fit normalization that
TABLE 1. The Main Clinicopathological Features and Results of Molecular Biologic Studies in 17 BSS/MFT Patients With Lack of a Demonstrable CYLD Germline Mutation Family Number 1 2 3 4 5 6 7
8 9 10 11 12 13
Patient Sample Sex/ Nationality/Ethnic Source of aCGH Number Phenotype Age Background DNA Analysis 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
MFT MFT MFT BSS BSS BSS MFT MFT MFT MFT MFT BSS BSS MFT MFT BSS BSS
M/36 F/30 M/46 F/59 M/38 F/61 F/28 M/21 F/14 M/38 F/19 F/43 F/50 F/49 F/20 F/47 M/64
Russian Czech German Czech Czech Spanish Spanish USA Czech Czech Czech Czech Belorussian Austrian German Czech English
B B FFPE B B FFPE B FFPE B B B FFPE B B B B FFPE
P P NA P NP NP P P NP P P NA P P NA P P
aCGH Result
Mutation Description
Deletion c.-34111_*297858del378779 Negative — — — Negative — — — — — Negative — Negative — — — Negative — Negative — — — Negative — Negative — — — Deletion c.914-6398_1769del13642ins20 Negative —
Number of Family/Patient in Grossmann et al, 2013 37/52 38/53 39/54 40/55 40/56 41/57 41/58 42/59 43/60 43/61 43/62 44/63 45/64 46/65 47/66 48/67 New case
B, blood; FFPE, formalin-fixed paraffin-embedded tissue; NP, not performed; NA, not analyzable; P, performed.
2014 Lippincott Williams & Wilkins
www.amjdermatopathology.com |
869
Vanecek et al
Am J Dermatopathol Volume 36, Number 11, November 2014
exported into .gpr, .txt, and .pdf files to allow further viewing and analysis of data in SignalMap (Roche NimbleGen Inc), general text editor or tab processor. Data were annotated in respect to the reference genome hg19 (Genome Reference Consortium Human Build 37).
Rearrangement Spanning PCR For exact determination of rearrangements, primers surrounding the affected region were designed for each suspect case (see Results and Table 2). All primers were designed using the Primer 3 software (http://frodo.wi.mit. edu/cgi-bin/primer3/primer3_www.cgi)46 and human chromosome 16 sequence (version GRCh37/hg19). Then, 100 ng of DNA was added to a mixture consisting of 10 pmole of forward and reverse primers (Table 2), 12.5 mL HotStar Taq PCR Master Mix (QIAgen, Hilden, Germany), and distilled water up to 25 mL. Initial denaturation at 958C for 14 minutes was followed by 40 cycles of denaturation at 958C for 1 minute, annealing at 608C for 1 minute, and extension at 728C for 2.5 minutes. The program was terminated by incubation at 728C for 10 minutes. The reactions were analyzed with agarose (Serva) gel electrophoresis. Successfully amplified PCR products were purified with magnetic beads using Agencourt AMPure kit (Agencourt Bioscience Corporation, A Beckman Coulter Company, Beverly, MA), both side sequenced using Big Dye Terminator Sequencing kit (Applied Biosystems, Carlsbad, CA), and purified with magnetic beads using Agencourt CleanSEQ kit (Agencourt Bioscience Corporation), all according to the manufacturer’s instructions. Samples were then run on an automated sequencer ABI Prism 3130xl (Applied Biosystems) at a constant voltage of 13.2 kV for 20 minutes. Data were processed in Sequencing Analysis software 5.3.1 (Applied Biosystems), and received sequences were compared with chromosome 16 reference sequence using BLAST software (http://blast.ncbi.nlm.nih.gov/Blast.cgi). Nucleotides from the targeted region were numbered using CYLD coding DNA reference sequence (NG_012061.1), and the ascertained mutations were described using current mutation nomenclature (http://www.hgvs.org/mutnomen/recs.html). FIGURE 1. Case 1. Clinicopathological features of the patient with MFTs. Multiple cutaneous nodules on the face (A) histopathologically representing trichoepitheliomas [cribriform trichoblastoma (B, C)].
compensates for inherent differences in signal between the 2 dyes was applied. The segMNT algorithm identifies copy number changes using a dynamic programming process that minimizes the squared error relative to the segment means. This procedure allows one to generate a list of candidate breakpoints, to identify the best segmentation for each given number of and to determine the number of segments to output results. The minimum segment difference in the log2 ratio that 2 segments must exhibit before they are identified as separate segments was set to 0.1. The minimum segment length was set to 2 probes. The stringency with which the initial segment boundaries were selected was set to the maximal stringent value 0.9. The segmentation nonaveraged and reduced averaging window was applied to raw data, which gave us data spacing of 100 and 100,000 bp, respectively. The results were
870
| www.amjdermatopathology.com
RESULTS First, we analyzed quantity, purity, and integrity of the DNA extracted from the patient’s peripheral blood and nontumor tissue samples using spectrophotometry and PCR, respectively. Eleven of the 14 specimens (9 extracted from blood, 2 extracted from formalin-fixed paraffin-embedded tissue) met the set criteria (Table 1). These 11 specimens were analyzed by high-resolution aCGH of chromosome 16. The aCGH analysis revealed a higher signal in the reference channel (red) comparing with a test channel (green) in a CYLD gene region, indicating a large deletion in 2 cases, namely cases 1 and 16 (Fig. 3). In case 1, a decrease in signal intensity in green channel started at probe position 50, 749, 626 and ended at probe position 51, 128, 051 on chromosome 16. In case 16, a decrease in green channel intensity started at probe position 50, 802, 675 and ended at probe position 50, 816, 050 on chromosome 16. Several sets of primers with increasing 2014 Lippincott Williams & Wilkins
Am J Dermatopathol Volume 36, Number 11, November 2014
Brooke–Spiegler Syndrome
TABLE 2. Primers Used for Rearrangement Spanning PCR Case
Name
Case 1 primer pairs
F1 R1 F2 R2 F3 R3 F3 R5 F6 R3 F6 R5 F8 R3 F8 R5 F10 R10 F1 R1 F3 R3 F4 R4 F4 R5 F6 R6 F6 R7 F8 R8 F2 R7 F8 R7
Case 16 primer pairs
Sequence 50 -30 GTCACTTCGCTTGTCTGTGC GCCTTCCATTGTGCATGATT GGCCATAAGAGGCCCTGAG AAGATGAAATGCCTCAAATTCC TGTTTCCATGCCAACTGAAA CTCAGGGAACCGTGAATGTT TGTTTCCATGCCAACTGAAA AAGAAGGGGGTTTTCCTCAA CAACCAGCTTGGTTTTCCAT CTCAGGGAACCGTGAATGTT CAACCAGCTTGGTTTTCCAT AAGAAGGGGGTTTTCCTCAA TCCTCATAGGCCAGTCCATC CTCAGGGAACCGTGAATGTT TCCTCATAGGCCAGTCCATC AAGAAGGGGGTTTTCCTCAA TGAGTGGGAATTTGGAAACC CCCCCAGGTTTGTGTCTTTA TGGCTTCTGAAGGATACCTTTG GCCTCCAAATGCTGAAAAGA CATCTGCCTTTCATGTCATCAT GGTATGTGTCCCTGCCTCAT TGGCTCAGAGCAGTGAAGAG CACAGAAACATGAGACAACATTCA TGGCTCAGAGCAGTGAAGAG TGCATATATGAGGGCAGCAG CGTCACTTGTCAGGCACACT GGCCGCCACTTTATTCTTCT CGTCACTTGTCAGGCACACT CTATTCAGTGCCCCACTTCC TTGAAGTGCTTTCTCCCACA AATTAGCCAGGCATGATGGA TTGCCCAATTACCTTCATCC CTATTCAGTGCCCCACTTCC TTGAAGTGCTTTCTCCCACA CTATTCAGTGCCCCACTTCC
FIGURE 2. Case 16. Clinicopathological features of the patient with BSS. Multiple cutaneous nodules on the forehead (A) histopathologically representing spiradenocylindromas (B, C).
on agarose gel electrophoresis. The lengths of products were of an approximately expected size (Fig. 4). All PCR fragments were then both side sequenced, and the exact positions of the rearrangements were defined. In case 1, a mutation c.34111_*297858del378779 (g.50749499_51128277del378779) was revealed. This deletion starts in intron 4 of the adjacent gene NOD2 and ends approximately 300 kB downstream of the stop codon in the CYLD gene.47 In case 16, a mutation c.914-6398_1769del13642ins20 (g.50802679_50816320del13642ins20) was found. The latter deletion starts in intron 6, ends in exon 12, of the CYLD gene and is accompanied by a small insertion of 20 bp of unknown origin (searched by BLAST) (Fig. 5; Table 1).
amplification range surrounding each rearrangement position were designed for each case (Table 2), and both DNA samples were amplified. After PCR amplification, 9 of 9 fragments in case 1 and 8 of 9 fragments in case 16 were detected
We analyzed a cohort of patients with BSS/MFT with the previously nondetectable germline sequence mutation in
2014 Lippincott Williams & Wilkins
DISCUSSION
www.amjdermatopathology.com |
871
Vanecek et al
Am J Dermatopathol Volume 36, Number 11, November 2014
FIGURE 3. The results of aCGH. A, The position of NOD2, CYLD, and SALL1 on chromosme 16. B, An approximately 0.4 MB deletion in case 1. C, An approximately 13.6 kB deletion in case 16.
coding region of the CYLD gene36 to search for large rearrangements, using aCGH followed by sequencing of rearrangement spanning PCR products in positive samples. Although a somatic mutation such as loss of heterozygosity caused by, among other things, a large deletion is apparently a common feature in patients with BSS/MFT,26,48,49,36 a germline alteration represented by a large deletion seems to be rare. To date, only 1 patient with BSS (reported as having familial cylindromatosis) with a large germline rearrangement in the CYLD gene has been described.42 This was an approximately 5.3-kB deletion involving and extending beyond the 30 end of the gene, affecting the USP domain of the CYLD protein. This patient was identified among 13 cases of BSS included in that study.42,3 In our study group, we found 2 novel large germline deletions in 2 patients, 1 with classical BSS phenotype and 1 with
MFT, representing 18% (2 of 14) of CYLD sequence mutation– negative analyzable cases and 3% (2 of 68) of the cohort. One of the 2 deletions was accompanied by a tiny insertion, thus technically representing a delins mutation. When the clinical phenotype is taken into account, large germline deletions are found in 2% and 5% of patients with classic BSS and MFT, respectively. No other large rearrangements (eg, duplication, insertion, and unbalanced translocation) were found in our cohort. The frequency of large rearrangements seems to be slightly lower when compared with other tumor suppressor genes affected in syndromes manifested by skin tumors, including Lynch syndrome/Muir–Torre syndrome and neurofibromatosis type 1 but comparable with tuberous sclerosis, Gorlin–Goltz, and Cowden syndromes.50–56 In our cohort, there is still a relatively high percentage of patients without detectable germline
FIGURE 4. The results of rearrangement spanning PCR. A, PCR products of expected sizes detected in 9 of 9 reactions in case 1. B, PCR products of expected sizes detected in 8 of 9 reactions in case 16. Normal control DNA amplified with designed primer sets and nontemplate controls were included but not shown. All control reaction were negative.
872
| www.amjdermatopathology.com
2014 Lippincott Williams & Wilkins
Am J Dermatopathol Volume 36, Number 11, November 2014
Brooke–Spiegler Syndrome
FIGURE 5. The results of sequencing analysis of selected PCR products. A, Arrow shows the position of the mutation in case 1. B, Arrows show the position of the mutation in case 16. Twenty base pair insert is also depicted.
mutation in the CYLD gene. Mutations in the coding region36 and large rearrangements have been excluded leaving alterations in intronic41 and regulatory regions, balanced translocations and, perhaps, epigenetic silencing, as possible explanations for BSS/MFT in our cohort. Less likely considerations for the pathogenesis of some BSS/MFT cases include modification at the expression level (eg, through microRNA) or another gene or regulatory pathway. The large molecular alterations found in 2 of our cases are quite different. The deletion (or technically delins mutation) in case 16 alters only approximately 13.5 kB of the CYLD gene, including the region encoding the third cytoskeleton-associated glycine-rich domain and the beginning of the USP domain. The mutation in case 1 is much more extensive, amounting to near complete deletion of the CYLD gene and the surrounding regions (including regulatory), affecting also a part of NOD2. The NOD2 gene is a member of the Nod1/Apaf-1 family and encodes a protein with 2 caspase recruitment domains and 6 leucine-rich repeats.57 NOD2 plays a role in the immune response, activating NF-kB signaling pathway, after recognition of intracellular bacterial lipopolysaccharides. Mutations in NOD2 have been associated with Crohn disease and Blau syndrome.58 No information was available to confirm or exclude the latter possibilities in our patient. In conclusion, we have identified 2 novel large germline rearrangement mutations including 1 pure deletion and 1 deletion/insertion (delins mutation) in a cohort of BSS/MFT. We have also demonstrated that large rearrangements are relatively rare in these patients. Other pathogenetic mechanisms in BSS/MFT, lacking germline sequence alterations, or large rearrangements in the CYLD gene, remain to be clarified. REFERENCES 1. Biggs PJ, Wooster R, Ford D, et al. Familial cylindromatosis (turban tumour syndrome) gene localised to chromosome 16q12-q13: evidence for its role as a tumour suppressor gene. Nat Genet. 1995;11:441–443.
2014 Lippincott Williams & Wilkins
2. Kazakov DV, Schaller J, Vanecek T, et al. Brooke-Spiegler syndrome: report of a case with a novel mutation in the CYLD gene and different types of somatic mutations in benign and malignant tumors. J Cutan Pathol. 2010;37:886–890. 3. Blake PW, Toro JR. Update of cylindromatosis gene (CYLD) mutations in Brooke-Spiegler syndrome: novel insights into the role of deubiquitination in cell signaling. Hum Mutat. 2009;30:1025–1036. 4. Kazakov DV, Soukup R, Mukensnabl P, et al. Brooke-Spiegler syndrome: report of a case with combined lesions containing cylindromatous, spiradenomatous, trichoblastomatous, and sebaceous differentiation. Am J Dermatopathol. 2005;27:27–33. 5. Bowen S, Gill M, Lee DA, et al. Mutations in the CYLD gene in BrookeSpiegler syndrome, familial cylindromatosis, and multiple familial trichoepithelioma: lack of genotype-phenotype correlation. J Invest Dermatol. 2005;124:919–920. 6. Ponti G, Nasti S, Losi L, et al. Brooke-Spiegler syndrome: report of two cases not associated with a mutation in the CYLD and PTCH tumorsuppressor genes. J Cutan Pathol. 2012;39:366–371. 7. Kazakov DV, Michal M, Kacerovska D, et al. Cutaneous adnexal tumors. Philadelphia, PA: Lippincott Williams & Wilkins; p.814. 2012. 8. Weyers W, Nilles M, Eckert F, et al. Spiradenomas in Brooke-Spiegler syndrome. Am J Dermatopathol. 1993;15:156–161. 9. Kazakov DV, Vanecek T, Zelger B, et al. Multiple (familial) trichoepitheliomas: a clinicopathological and molecular biological study, including CYLD and PTCH gene analysis, of a series of 16 patients. Am J Dermatopathol. 2011;33:251–265. 10. Clarke J, Ioffreda M, Helm KF. Multiple familial trichoepitheliomas: a folliculosebaceous-apocrine genodermatosis. Am J Dermatopathol. 2002;24:402–405. 11. Puig L, Nadal C, Fernandez-Figueras MT, et al. Brooke-Spiegler syndrome variant: segregation of tumor types with mixed differentiation in two generations. Am J Dermatopathol. 1998;20:56–60. 12. Petersson F, Kutzner H, Spagnolo DV, et al. Adenoid cystic carcinomalike pattern in spiradenoma and spiradenocylindroma: a rare feature in sporadic neoplasms and those associated with Brooke-Spiegler syndrome. Am J Dermatopathol. 2009;31:642–648. 13. Kazakov DV, Vanecek T, Nemcova J, et al. Spectrum of tumors with follicular differentiation in a patient with the clinical phenotype of multiple familial trichoepitheliomas: a clinicopathological and molecular biological study, including analysis of the CYLD and PTCH genes. Am J Dermatopathol. 2009;31:819–827. 14. Uede K, Yamamoto Y, Furukawa F. Brooke-Spiegler syndrome associated with cylindroma, trichoepithelioma, spiradenoma, and syringoma. J Dermatol. 2004;31:32–38. 15. Pizinger K, Michal M. Malignant cylindroma in Brooke-Spiegler syndrome. Dermatology. 2000;201:255–257.
www.amjdermatopathology.com |
873
Vanecek et al
16. Kazakov DV, Zelger B, Rutten A, et al. Morphologic diversity of malignant neoplasms arising in preexisting spiradenoma, cylindroma, and spiradenocylindroma based on the study of 24 cases, sporadic or occurring in the setting of Brooke-Spiegler syndrome. Am J Surg Pathol. 2009;33: 705–719. 17. De Francesco V, Frattasio A, Pillon B, et al. Carcinosarcoma arising in a patient with multiple cylindromas. Am J Dermatopathol. 2005;27:21–26. 18. Antonescu CR, Terzakis JA. Multiple malignant cylindromas of skin in association with basal cell adenocarcinoma with adenoid cystic features of minor salivary gland. J Cutan Pathol. 1997;24:449–453. 19. Pincus LB, McCalmont TH, Neuhaus IM, et al. Basal cell carcinomas arising within multiple trichoepitheliomas. J Cutan Pathol. 2008;35 (suppl 1):59–64. 20. Pingitore R, Campani D. Salivary gland involvement in a case of dermal eccrine cylindroma of the scalp (turban tumor). Report of a case with lung metastases. Tumori. 1984;70:385–388. 21. Kakagia D, Alexiadis G, Kiziridou A, et al. Brooke-Spiegler syndrome with parotid gland involvement. Eur J Dermatol. 2004;14:139–141. 22. Kostler E, Schonlebe J, Mentzel T, et al. Psoriasis and Brooke-Spiegler syndrome with multiple malignancies. J Eur Acad Dermatol Venereol. 2005;19:380–381. 23. Scott AR, Faquin WC, Deschler DG. Parotid mass in a woman with multiple cutaneous cylindromas. Head Neck. 2010;32:684–687. 24. Nonaka D, Rosai J, Spagnolo D, et al. Cylindroma of the breast of skin adnexal type: a study of 4 cases. Am J Surg Pathol. 2004;28:1070–1075. 25. Kazakov DV, Spagnolo DV, Kacerovska D, et al. Cutaneous type adnexal tumors outside the skin. Am J Dermatopathol. 2011;33:303–315. 26. Bignell GR, Warren W, Seal S, et al. Identification of the familial cylindromatosis tumour-suppressor gene. Nat Genet. 2000;25:160–165. 27. Thomson SA, Rasmussen SA, Zhang J, et al. A new hereditary cylindromatosis family associated with CYLD1 on chromosome 16. Hum Genet. 1999;105:171–173. 28. Lv HL, Huang YJ, Zhou D, et al. A novel missense mutation of CYLD gene in a Chinese family with multiple familial trichoepithelioma. J Dermatol Sci. 2008;50:143–146. 29. Oiso N, Mizuno N, Fukai K, et al. Mild phenotype of familial cylindromatosis associated with an R758X nonsense mutation in the CYLD tumour suppressor gene. Br J Dermatol. 2004;151:1084–1086. 30. Biggs PJ, Chapman P, Lakhani SR, et al. The cylindromatosis gene (cyld1) on chromosome 16q may be the only tumour suppressor gene involved in the development of cylindromas. Oncogene. 1996;12: 1375–1377. 31. Massoumi R, Paus R. Cylindromatosis and the CYLD gene: new lessons on the molecular principles of epithelial growth control. Bioessays. 2007; 29:1203–1214. 32. Masoumi KC, Shaw-Hallgren G, Massoumi R. Tumor suppressor function of CYLD in nonmelanoma skin cancer. J Skin Cancer. 2011;2011: 614097. 33. Tauriello DV, Haegebarth A, Kuper I, et al. Loss of the tumor suppressor CYLD enhances Wnt/beta-catenin signaling through K63-linked ubiquitination of Dvl. Mol Cell. 2010;37:607–619. 34. Young AL, Kellermayer R, Szigeti R, et al. CYLD mutations underlie Brooke-Spiegler, familial cylindromatosis, and multiple familial trichoepithelioma syndromes. Clin Genet. 2006;70:246–249. 35. Wickstrom SA, Masoumi KC, Khochbin S, et al. CYLD negatively regulates cell-cycle progression by inactivating HDAC6 and increasing the levels of acetylated tubulin. EMBO J. 2010;29:131–144. 36. Grossmann P, Vanecek T, Steiner P, et al. Novel and recurrent germline and somatic mutations in a cohort of 67 patients from 48 families with Brooke-Spiegler syndrome including the phenotypic variant of multiple familial trichoepitheliomas and correlation with the histopathologic findings in 379 biopsy specimens. Am J Dermatopathol. 2013;35:34–44. 37. Almeida S, Maillard C, Itin P, et al. Five new CYLD mutations in skin appendage tumors and evidence that aspartic acid 681 in CYLD is essential for deubiquitinase activity. J Invest Dermatol. 2008;128:587–593.
874
| www.amjdermatopathology.com
Am J Dermatopathol Volume 36, Number 11, November 2014
38. Kacerovska D, Szep Z, Kollarikova L, et al. A novel germline mutation in the CYLD gene in a Slovak patient with Brooke-Spiegler syndrome. Cesk Patol. 2013;49:89–92. 39. Zheng G, Hu L, Huang W, et al. CYLD mutation causes multiple familial trichoepithelioma in three Chinese families. Hum Mutat. 2004;23:400. 40. Linos K, Schwartz J, Kazakov DV, et al. Recurrent CYLD nonsense mutation associated with a severe, disfiguring phenotype in an African American family with multiple familial trichoepithelioma. Am J Dermatopathol. 2011;33:640–642. 41. Kazakov DV, Thoma-Uszynski S, Vanecek T, et al. A case of BrookeSpiegler syndrome with a novel germline deep intronic mutation in the CYLD gene leading to intronic exonization, diverse somatic mutations, and unusual histology. Am J Dermatopathol. 2009;31:664–673. 42. van den Ouweland AM, Elfferich P, Lamping R, et al. Identification of a large rearrangement in CYLD as a cause of familial cylindromatosis. Fam Cancer. 2010;10:127–132. 43. Saggar S, Chernoff KA, Lodha S, et al. CYLD mutations in familial skin appendage tumours. J Med Genet. 2008;45:298–302. 44. van Dongen JJ, Langerak AW, Bruggemann M, et al. Design and standardization of PCR primers and protocols for detection of clonal immunoglobulin and T-cell receptor gene recombinations in suspect lymphoproliferations: report of the BIOMED-2 Concerted Action BMH4-CT98-3936. Leukemia. 2003;17:2257–2317. 45. Workman C, Jensen LJ, Jarmer H, et al. A new non-linear normalization method for reducing variability in DNA microarray experiments. Genome Biol. 2002;3:research0048. 46. Rozen S, Skaletsky H. Primer3 on the WWW for general users and for biologist programmers. Methods Mol Biol. 2000;132:365–386. 47. Moreira LO, Zamboni DS. NOD1 and NOD2 signaling in infection and inflammation. Front Immunol. 2012;3:328. 48. Leonard N, Chaggar R, Jones C, et al. Loss of heterozygosity at cylindromatosis gene locus, CYLD, in sporadic skin adnexal tumours. J Clin Pathol. 2001;54:689–692. 49. Sima R, Vanecek T, Kacerovska D, et al. Brooke-Spiegler syndrome: report of 10 patients from 8 families with novel germline mutations: evidence of diverse somatic mutations in the same patient regardless of tumor type. Diagn Mol Pathol. 2010;19:83–91. 50. Peltomaki P, Vasen H. Mutations associated with HNPCC predisposition— update of ICG-HNPCC/INSiGHT mutation database. Dis Markers. 2004; 20:269–276. 51. Strizheva GD, Carsillo T, Kruger WD, et al. The spectrum of mutations in TSC1 and TSC2 in women with tuberous sclerosis and lymphangiomyomatosis. Am J Respir Crit Care Med. 2001;163:253–258. 52. Brook-Carter PT, Peral B, Ward CJ, et al. Deletion of the TSC2 and PKD1 genes associated with severe infantile polycystic kidney disease— a contiguous gene syndrome. Nat Genet. 1994;8:328–332. 53. Kacerovska D, Vrtel R, Michal M, et al. TSC2/PKD1 contiguous gene syndrome: a report of 2 cases with emphasis on dermatopathologic findings. Am J Dermatopathol. 2009;31:532–541. 54. Longa L, Scolari F, Brusco A, et al. A large TSC2 and PKD1 gene deletion is associated with renal and extrarenal signs of autosomal dominant polycystic kidney disease. Nephrol Dial Transplant. 1997; 12:1900–1907. 55. Eng C. The role of PTEN, a phosphatase gene, in inherited and sporadic nonmedullary thyroid tumors. Recent Prog Horm Res. 1999;54:441–452; discussion 453. 56. Marsh DJ, Coulon V, Lunetta KL, et al. Mutation spectrum and genotype-phenotype analyses in Cowden disease and BannayanZonana syndrome, two hamartoma syndromes with germline PTEN mutation. Hum Mol Genet. 1998;7:507–515. 57. Hugot JP, Chamaillard M, Zouali H, et al. Association of NOD2 leucinerich repeat variants with susceptibility to Crohn’s disease. Nature. 2001; 411:599–603. 58. Abraham C, Cho JH. Functional consequences of NOD2 (CARD15) mutations. Inflamm Bowel Dis. 2006;12:641–650.
2014 Lippincott Williams & Wilkins
