Somatic mutation in exon 6 of the 3β-hydroxysteroid dehydrogenase (NSDHL) gene in verruciform xanthoma. The second base of codon 199 (CGC) of the NSDHL gene is guanosine (arrow) in normal control (A), and this is converted to adenosine (arrow) in one of the verriuciform xanthomas (case 4) (B).
Mehra S, Li L, Fan C, Smoller B, Morgan M, Somach S. A Novel Somatic Mutation of the 3β-Hydroxysteroid Dehydrogenase Gene in Sporadic Cutaneous Verruciform Xanthoma. Arch Dermatol. 2005;141(10):1263-1267. doi:10.1001/archderm.141.10.1263
To analyze the 3β-hydroxysteroid dehydrogenase (NSDHL) gene in verruciform xanthoma (VX) to elucidate its potential role in the histogenesis of this lesion.
DNA was extracted from paraffin-embedded tissue, followed by polymerase chain reaction amplification of exons 4 and 6 of the NSDHL gene. The polymerase chain reaction products were then directly sequenced and analyzed for the presence of somatic mutations.
Nine lesions of VX from 8 patients and 3 unrelated normal controls were evaluated.
Two of 9 VXs (22%) demonstrated a novel somatic missense mutation in exon 6 of the NSDHL gene. The mutation was not present in the remaining 7 lesions of VX, nonlesional internal controls, and 3 unrelated normal controls. No mutation of exon 4 was found in any case. Mutations of exons 4 and 6 previously identified in CHILD syndrome were not seen in our cases.
(1) A novel missense mutation (R199H) in exon 6 of the NSDHL gene was identified in a small subset of sporadic VXs. (2) Known CHILD syndrome mutations in exons 4 and 6 of the NSDHL gene do not contribute to the histogenesis of sporadic VXs.
Verruciform xanthoma (VX) is a rare mucocutaneous lesion of uncertain etiology. First described by Shafer in 1971,1 it was originally thought to be limited to oral mucosa. Subsequent studies have shown its occurrence in other mucosal and periorificial areas as well as nonmucosal areas. Verruciform xanthoma has been associated with a wide variety of disorders such as recessive dystrophic epidermolysis bullosa,2 lymphedema,3 graft vs host disease, and epidermal nevus.4,5 It has also been reported in patients with no other associated abnormality. The occurrence of VX in other disorders has led to the suggestion that it is a morphologic change rather than a true entity.6,7 A viral etiology has been suggested and studied widely because of the lesion’s architecture and occurrence in genital and oral areas.8- 11
The histologic characterization of VX is verrucous epithelial hyperplasia of squamous epithelium with crypts of premature keratinization and collections of lipid-laden macrophages within the superficial papillary dermis or submucosa. Identical morphologic changes are observed in cutaneous lesions of CHILD syndrome (congenital hemidysplasia, ichthyosiform nevus, and limb defects). CHILD syndrome, an X-linked dominant male-lethal trait, is caused by mutational inactivation of the 3β-hydroxysteroid dehydrogenase (NSDHL) gene, located at Xq28.8 The NSDHL gene plays an important role in cholesterol biosynthesis by encoding for the NSDHL enzyme that is required for cholesterol biosynthesis. Because the skin lesions found in CHILD syndrome closely resemble sporadic cutaneous VX, we hypothesized that inactivation of the NSDHL gene may also occur in sporadic VX. In this study, we performed mutational analysis of exons 4 and 6 of the NSDHL gene in sporadic VX cases to elucidate a potential role of NSDHL in the histogenesis of this lesion. We elected to investigate exons 4 and 6 only because 6 of 9 reported cases of CHILD syndrome with mutations in the NSDHL gene showed mutations in these 2 exons (Table 1).
Nine cases of cutaneous VX, with 3 unrelated normal controls and 2 internal controls (2b and 7c from cases 2 and 7, respectively) (Table 2), were evaluated. Patient 7 had 2 different lesions (7a and 7b) (Table 2), both on the scrotum but appearing at different times during the same year. Paraffin blocks were retrieved from the archives (1990-2002) of the Department of Pathology, MetroHealth Medical Center, Case Western Reserve University School of Medicine, Cleveland, Ohio, and the Cleveland Skin Pathology Laboratory. Paraffin blocks and hematoxylin-eosin–stained slides of each case were available for study. Three randomly selected skin biopsy specimens with no specific pathologic diagnosis were used for DNA isolation as a negative control.
Five tissue sections (5 μm thick) from each selected paraffin block were deparaffinized by 2 washes in 100% xylene followed by 2 washes in 100% ethanol. The deparaffinized tissues were then subjected to DNA extraction using the EX-WAX DNA Extraction Kit (Intergen Co, New York, NY), according to manufacturer’s instruction.
Primer sets were designed to amplify mutational hot spots in CHILD syndrome within exons 4 and 6 of the NSDHL gene. The primer sequences are 5′-CCA GCT CTG AAA GGT GTA AAC ACA-3′ (sense) and 5′-CAA GTT TCA ATG ACA TTC TTG GTG CC-3′ (antisense) for exon 4; 5′-A GTT CTG GGC GCC AAC GAT-3′ (sense) and 5′-CAA TCA CGA ACT TCA TCT TGC CGT-3′ (antisense) for exon 6. The expected size of polymerase chain reaction (PCR) amplicons is 119 base pairs for exon 4 and 140 base pairs for exon 6. Amplification of NSDHL exons 4 and 6 was carried out in a Touchgene Gradient Thermal Cycler (Techne Inc, Princeton, NJ) in a 50-μL PCR mixture containing 2 μL of genomic DNA, dNTPs (200μM of each), primers (50pM each per reaction), 2.5mM magnesium chloride, and 1.25 U Hotstar Taq (Qiagen, Valencia, Calif) in 1× PCR buffer. All reagents are supplied with the Hotstar Taq Kit (Qiagen) except for the dNTP mix (Roche Molecular Biochemicals, Indianapolis, Ind). Thermocycling conditions used were initial denaturation and hot start at 95°C for 15 minutes, 40 cycles consisting of 30 seconds at 95°C, 30 seconds at 58°C, and then 1 minute at 72°C. Following thermocycling, reactions were subjected to a 5-minute 72°C incubation. Polymerase chain reaction amplicons were visualized by electrophoresis on 2% agarose gels and visualized by ethidium bromide staining.
Amplified PCR products were ligated into a pCR4-TOPO vector (Invitrogen Life Technologies, Carlsbad, Calif) and transformed into Escherichia coli using a TOPO TA cloning kit (Invitrogen Life Technologies) for sequencing. Plasmid DNA isolated from E coli colonies was sequenced using a DNA sequencer (model 377; Applied Biosystems, Foster City, Calif) as previously described.12 The templates used for the sequencing were at the concentration of 100 ng/μL. A T3 universal primer was used in all reactions at 1.6μM. A Dye Terminator Kit (version 2.0; Applied Biosystems) was used according to manufacturer's instructions.
Most patients were seen during their eighth or ninth decades. Lesions were found equally among men and women. The clinicopathological characteristics of 8 patients with VX are summarized in Table 3. The most common initial clinical impression was that of polyp or squamous cell carcinoma. None of these patients had a personal or family history of CHILD syndrome.
Two of 9 VX lesions (22%) demonstrated a novel somatic mutation in exon 6 of the NSDHL gene. This mutation in exon 6 was found in cases 4 and 7 (lesion 7a). This is a missense mutation that occurs at nucleotide 596 (G596A; G→A transition mutation) of the NSDHL gene (Figure). This mutation occurred at the second base of codon 199 (CGC→CAC) and resulted in the replacement of arginine by histidine at the codon (R199H) in the final protein product. No somatic mutation was found in exon 4 of the NSDHL gene in any of the lesions. The R199H missense mutation was not detected in the remaining 7 lesions of VX, a nonlesional internal control from cases 2 (2b) and 7 (7c), and 3 unrelated normal controls (cases 9-11) (Table 2). Controls did not show any mutation in exon 4 or 6. None of the mutations that have been previously identified on exons 4 and 6 in CHILD syndrome (A105V of exon 48; A182P, G205S, and Q210X of exon 69) were seen in VX cases. Table 2 gives a summary of NSDHL mutational analysis of all VX cases.
Verruciform xanthoma is a histopathologic pattern seen in various cutaneous or mucosal lesions of heterogeneous etiology. Verrucous epithelial hyperplasia of squamous epithelium with aggregates of lipid-laden macrophages in the submucosa or papillary dermis is the hallmark of this lesion. Common locations include oral mucosa and anogenital and periorificial skin. A human papilloma virus (HPV) etiology has been suggested and studied widely.13- 16 Other suggested possible causes include reaction to microorganisms, based on the abundance of neutrophils within premature keratinization.1,17- 20 However, no organisms have been identified.1,10,11,17,18
Evaluation for HPV in VX has been studied and has included types 6, 7, 11, 12, 16 to 19, 31, 33, 35, 37, 44 to 47, 53, 54, and 58.6,13,21- 23 While most studies failed to show evidence of HPV, 1 study13 demonstrated by immunohistochemical analysis a positive reaction in the upper epidermal keratinocytes, which was confirmed by electron microscopy and by PCR, which also detected HPV-DNA 6a. Another investigation22 demonstrated by immunohistochemical analysis a positive reaction with polyclonal antisera against papilloma virus. This reaction was observed in histiocytes and could not be confirmed by in situ hybridization. Thus, the authors concluded that this reaction could be the result of cross-reactivity to HPV-unrelated antigens. One more positive reaction in histiocytes with HPV antisera,6 which could not be confirmed by electron microscopy, has also been reported. Similarly, Iamaroon and Vickers21 demonstrated a positive reaction with anti-HPV antibodies by in situ hybridization in 1 of 12 cases. However, this study also failed to confirm the positive reaction by immunohistochemical analysis. Thus, so far no conclusive viral etiology has been found.
Although studies have identified the foamy cells as lipid-laden macrophages,5,14,16,18 the process by which the lipid accumulates has not been elucidated. Zegarelli et al18 proposed that epithelial degeneration, along with loss of basal lamina, probably caused by a local irritant, is the initiating event. This supports the theory that a degenerating epidermis is the source of lipid for the dermal histiocytes.22,24 However, Travis et al16 suggested that the accumulation of foam cells is the primary abnormality and that the epithelial hyperplasia and inflammation are secondary. Cases of VX occurring with lymphedema support this theory.3
Cholesterol is an essential component of cell membranes and plays an important role in membrane fluidity. The higher the concentration of cholesterol, the thicker and less permeable the membrane is. Cholesterol concentration in different cellular membranes depends on functional needs of the organelle. Genes involved in late cholesterol biosynthesis have been identified only recently; NSDHL is one of these genes.25
The NSDHL gene plays an important role in cholesterol biosynthesis pathway because it encodes NSDHL. A total of 8 exons of this gene have been described. NSDHL is localized on the surface of lipid storage droplets (LDs), which are thought to originate from the endoplasmic reticulum (ER). NSDHL localization on LDs is very specific and is highly conserved throughout its evolution from yeast to mammals. Ohashi et al26 showed that the presence of well-developed LDs with NSDHL acts as a rate-limiting step and reduces the formation of cholesterol by negative feedback. Formation of LDs also depends on the concentration of fatty acids. Depletion of fatty acids leads to reduced formation of LDs and thus redistribution of NSDHL to the ER. Ohashi et al26 also showed that human NSDHL with a nonsense mutation causing CHILD syndrome could no longer be localized on the surface of LDs. Mutated NSDHL failed to restore the defective growth of Chinese hamster ovarian cell cholesterol auxotroph, even in a cholesterol-deficient medium. These findings suggest the functional significance of NSDHL localization on the LD surface. Lipid storage droplets traditionally have been considered as depots for lipids including cholesterol and cholesteryl esters.27 However, the findings described by Ohashi et al26 suggest that the LD is not simply a cholesterol depot but may also be involved in regulation of cholesterol biosynthesis. Thus, the findings of Ohashi et al26 suggest that once enough LDs are formed, depending on the body needs and availability of fatty acids, NSDHL moves from ER to LDs. This sends a negative feedback to stop producing more LDs. However, once NSDHL is mutated it can no longer migrate from ER to LDs. This results in loss of negative feedback and could theoretically cause excess formation and accumulation of LDs, which leads to subepithelial lipid-laden macrophages as seen in VX.
Codon 199 is located in an area that is highly conserved in all NSDHLs among different species. Thus, the amino acid change at this codon (R199H) due to missense mutation (G596A) is likely to be functionally significant. Even though lesions 7a and 7b are from the same patient and have identical morphologic traits, only lesion 7a showed this mutation. This finding supports the theory that G596A is an acquired somatic mutation. The clinical and morphological parameters of these 2 lesions were indistinguishable. The presence of the G596A mutation in only 2 (22%) of 9 lesions further suggests the possibility of multifactorial etiology or the presence of additional mutations on other exons that have not been studied yet.
Recently, Konig et al8,9 reported 2 nonsense mutations (R88X on exon 3, Q210X on exon 6) and 3 missense mutations (A105V on exon 4, G205S and A182P on exon 6) in CHILD syndrome cases. Hummel et al10 reported nonsense mutation E151X on exon 5, and Murata et al11 have reported another mutation, Y349C on exon 8 of the NSDHL gene. Although none of the reported exon 4 and 6 mutations are detected in our VX cases, the occurrence of a mutation in the NSDHL gene, along with morphological similarities in the skin lesions of CHILD syndrome and VX, supports the functional role of this gene in the pathogenesis of VX. Because we did selective analysis of the exons expected to have the highest yield, we suspect that mutations may be present within other exons not assessed. Thus, additional studies of VX are needed to evaluate the entire NSDHL gene for other possible mutations.
Correspondence: Stephen Somach, MD, Hamann 520, MetroHealth Medical Center, 2500 MetroHealth Dr, Cleveland, OH 44109 (firstname.lastname@example.org).
Accepted for Publication: December 22, 2004.
Author Contributions: Drs Mehra and Li contributed equally to the work presented in this article.
Financial Disclosure: None.