Skip to main content
PMC home page
eLife logoLink to eLife
. 2015 Jul 23;4:e06322. doi: 10.7554/eLife.06322

Genomic variations of the mevalonate pathway in porokeratosis

Zhenghua Zhang 1,*,†, Caihua Li 2,3,†, Fei Wu 4, Ruixiao Ma 3, Jing Luan 1, Feng Yang 3, Weida Liu 5, Li Wang 3, Shoumin Zhang 6, Yan Liu 3, Jun Gu 7, Wenlian Hua 3, Min Fan 8, Hua Peng 3, Xuemei Meng 9, Ningjing Song 10, Xinling Bi 7, Chaoying Gu 1, Zhen Zhang 1, Qiong Huang 1, Lianjun Chen 1, Leihong Xiang 1, Jinhua Xu 1, Zhizhong Zheng 1, Zhengwen Jiang 3,11,*
Editor: Helen H Hobbs12
PMCID: PMC4511816  PMID: 26202976

Abstract

Porokeratosis (PK) is a heterogeneous group of keratinization disorders. No causal genes except MVK have been identified, even though the disease was linked to several genomic loci. Here, we performed massively parallel sequencing and exonic CNV screening of 12 isoprenoid genes in 134 index PK patients (61 familial and 73 sporadic) and identified causal mutations in three novel genes (PMVK, MVD, and FDPS) in addition to MVK in the mevalonate pathway. Allelic expression imbalance (AEI) assays were performed in 13 lesional tissues. At least one mutation in one of the four genes in the mevalonate pathway was found in 60 (98%) familial and 53 (73%) sporadic patients, which suggests that isoprenoid biosynthesis via the mevalonate pathway may play a role in the pathogenesis of PK. Significantly reduced expression of the wild allele was common in lesional tissues due to gene conversion or some other unknown mechanism. A G-to-A RNA editing was observed in one lesional tissue without AEI. In addition, we observed correlations between the mutations in the four mevalonate pathway genes and clinical manifestations in the PK patients, which might support a new and simplified classification of PK under the guidance of genetic testing.

DOI: http://dx.doi.org/10.7554/eLife.06322.001

Research organism: human

eLife digest

Porokeratosis refers to a group of around twenty skin conditions that involve a build-up of a protein called keratin in skin cells. Keratin forms the tough fibres that give strength to hair and nails, and people suffering from porokeratosis develop hardened skin lesions. Porokeratosis is an uncommon condition; most cases are inherited and have been linked to exposure to ultraviolet light and having a weakened immune system.

Mutations in one gene called MVK are known to cause two forms of the disorder, but it is suspected that other genetic causes of porokeratosis will also be identified. The MVK gene encodes an enzyme that is involved in making chemicals called isoprenoids. This large and diverse class of chemicals provides the building blocks for making many other important molecules in all living species. Zhang, Li et al. have now analysed genetic material from 134 different porokeratosis patients to search for mutations in other genes involved in the production of isoprenoids. The patients examined include 61 people with a family history of the disorder, and 73 cases in which the condition seems to be a one-off occurrence.

This search identified mutations in three additional genes (called PMVK, MVD and FDPS) that are all linked to porokeratosis. Further analysis of these three genes and MVK revealed that about half of the patients with mutations in the MVK gene developed large lesions (that were over 5 centimetres in diameter). However, those with mutations in the other three genes did not develop such large lesions. Mutations in some of the newly identified genes were instead linked to porokeratosis affecting specific areas of the body; for example, PMVK and MVD mutations are linked to porokeratosis localized to the genitals and around the eyes, respectively. This means that, in the future, doctors might be able to simplify the diagnosis of the different varieties of porokeratosis based on information gained via genetic tests.

DOI: http://dx.doi.org/10.7554/eLife.06322.002

Introduction

Porokeratosis (PK, MIM 175800) is a heterogeneous group of keratinization disorders that exhibit an autosomal dominant mode of inheritance. PK is also a skin-specific autoinflammatory disease which was often inherited and linked to ultraviolet light exposure and immunosuppression (Schamroth et al., 1997; Abramovits and Oquendo, 2013). For example, eruptive pruritic papular porokeratosis exemplifies the inflammatory manifestation, and complications to inflammatory conditions such as localized cutaneous amyloidosis are seen in PK patients (Biswas, 2015). Indeed, PK and psoriasis share some features at both clinical and molecular levels and sometimes coexist in the same patients (Zhang et al., 2008).

As a histological hallmark that unifies all variants of PK, cornoid lamella (CL) is a vertical ‘column’ of parakeratosis. The pattern of CL can be slender, broad, or confluent, which is related to epidermal hyperplasia and dermal inflammation. However, CL is not a unique feature of PK because it can be seen in some inflammatory and inherited cutaneous disorders and also as an incidental finding (Biswas, 2015).

PK is currently classified according to the clinical manifestations, such as number, size, morphology, and distribution of the histological lesions. A better system of classification is expected because some variants of PK are fraught with confusing terminology (Schamroth et al., 1997; Sertznig et al., 2012; Biswas, 2015). For example, it is sometimes hard to completely differentiate disseminated superficial actinic porokeratosis (DSAP) from disseminated superficial porokeratosis (DSP) by age of onset and sun-exposed areas.

In addition to the heterogeneity in clinical manifestations, genetic heterogeneity is also observed in PK. At least five linkage loci (i.e., 12q23.2-24.1, 15q25.1-26.1, 18p11.3, 1p31.3-p31.1, 16q24.1-24.3) have been reported for the disseminated forms of PK which include DSAP, DSP, porokeratosis palmaris et plantaris disseminata (PPPD), and immunosuppression-induced porokeratosis (ISIP) (Schamroth et al., 1997; Luan et al., 2011). However, only one causal gene, the mevalonate kinase gene (MVK) at 12q24, has been identified in DSAP and porokeratosis of Mibille (PM) (Zhang et al., 2012; Li and Zhang, 2014; Zeng et al., 2014). Here we performed genetic analysis in 134 index patients with PK to identify additional causal genes and to establish new parameters for better differential diagnosis of PK. Allelic expression imbalance (AEI) and cDNA mutation were analyzed in 13 pairwise lesional tissues (LTs) and neighboring normal-appearing skin (NNS) to investigate the underlying pathogenicity of PK.

Results

Co-segregation of mevalonate (diphospho) decarboxylase (MVD) mutation with PK

We previously identified a linkage locus for DSAP on chromosome 16q24.1-24.3 in a four-generation Chinese DSAP family (Luan et al., 2011). We applied the candidate gene approach and hypothesized that the MVD gene in the 16q24.1-24.3 region was the causal gene of DSAP in the PK family. We performed Sanger sequencing of all MVD exons and identified a c.746T>C (p.Phe249Ser) mutation in MVD. The c.746T>C mutation displayed 100% co-segregation with the PK phenotype in this family (Figure 1).

Figure 1. Identification of a MVD mutation in a porokeratosis (PK) family.

(A) c.746T>C (p.Phe249Ser) in MVD displayed 100% co-separation with PK phenotype in this family (Luan et al., 2011). (B) Sanger sequencing chromatograms of proband (II-7, affected) and normal control (II-8, unaffected) at the c.746T>C mutation site indicated by arrow.

DOI: http://dx.doi.org/10.7554/eLife.06322.003

Figure 1.

Figure 1—figure supplement 1. Examples of six pedigree charts showing that each mutation displayed 100% co-segregation with the porokeratosis (PK) phenotype in the family.

Figure 1—figure supplement 1.

DOI: http://dx.doi.org/10.7554/eLife.06322.004
Figure 1—figure supplement 2. Two MVD mutations, c.302C>G (p.Pro101Arg) and c.683 G>A (p.Arg228Gln), for S-62 were located in the trans position because his daughter (S-62-D) carried only one of them.

Figure 1—figure supplement 2.

DOI: http://dx.doi.org/10.7554/eLife.06322.005
Open in a new tab

Point mutations in MVK, PMVK, MVD, and FDPS were identified in PK patients

Since both MVD and MVK are involved in isoprenoid biosynthesis via the mevalonate pathway (Thurnher et al., 2013), we hypothesized that mutations in other members of the mevalonate pathway exist in PK and screened for mutations in 12 genes in the mevalonate pathway including MVD and MVK in 134 PK patients (61 familial and 73 sporadic). The 12 genes are: acetyl-CoA acetyltransferase (ACAT) 1 and 2, 3-hydroxy-3-methylglutaryl-CoA synthase (HMGCS) 1 and 2, 3-hydroxy-3-methylglutaryl-CoA reductase (HMGCR), MVK, phosphomevalonate kinase (PMVK), MVD, isopentenyl-diphosphate delta isomerase (IDI) 1 and 2, farnesyl diphosphate synthase (FDPS), and geranylgeranyl diphosphate synthase 1 (GGPS1) (Kanehisa et al., 2012) (Figure 2). A cycled primer extension and ligation-dependent amplification (CPELA) reaction for enrichment of target genomic fragments and next generation sequencing (NGS) were performed in each sample. The NGS data were analyzed for rare single nucleotide variants (SNVs) with no record in dbSNP or a minor allele frequency of <1% in Chinese from 1000 genome database. After Sanger sequencing validation, 60 out of 61 SNVs called from NGS data were demonstrated to be true SNVs, which were predicted to impact protein function. Of these, 12 non-pathogenic SNVs were excluded from subsequent analysis (Supplementary file 1). In total, 48 mutation sites were identified in MVK, PMVK, MVD, and FDPS (Figure 3, Figure 3—source data 1).

Figure 2. Isoprenoid biosynthesis via the mevalonate pathway.

Figure 2.

Open in a new tab

12 member genes (ACAT1, ACAT2, HMGCS1, HMGCS2, HMGCR, MVK, PMVK, MVD, IDI1, IDI2, FDPS, GGPS1) were subject to mutation screening. The genomic loci of the 12 member genes are provided in parentheses. The illustration is adapted from the 00900 interactive map of the Kyoto Encyclopedia of genes and genomes (KEGG) (Kanehisa et al., 2012).

DOI: http://dx.doi.org/10.7554/eLife.06322.006

Figure 3. Mutational spectrum of MVK, PMVK, MVD and FDPS in 113 of the 134 porokeratosis (PK) patients.

DOI: http://dx.doi.org/10.7554/eLife.06322.007

Figure 3—source data 1. Sanger sequencing chromatograms of normal control and porokeratosis (PK) patients at 48 mutation sites in MVK, PMVK, MVD and FDPS.
DOI: http://dx.doi.org/10.7554/eLife.06322.008
elife06322s001.pdf (966.6KB, pdf)
DOI: 10.7554/eLife.06322.008

Figure 3.

Figure 3—figure supplement 1. Breakpoint analysis for three large deletion mutations in MVK and FDPS genes.

Figure 3—figure supplement 1.

(A) Agarose electrophoresis analysis of long PCR products from patients with deletion mutations and a normal control. (B) Sequencing chromatograms of long PCR products from deletion mutations. (C) Alignment of the sequences from three deletion mutations with human genome reference GRch38 primary assembly indicates a deletion of 10,076 bp, 3362 bp and 3304 bp, respectively.
DOI: http://dx.doi.org/10.7554/eLife.06322.009
Figure 3—figure supplement 2. Illustration of amplification of multiple target DNA fragments mediated by cycled primer extension and ligation.

Figure 3—figure supplement 2.

In brief, for each target region, an extension primer and a block probe are designed. The extension primer has a 5′ exonuclease-resistance modification and the block probe has a 5′ phosphorate and 3′exonuclease-resistance modification. These two oligos are mixed with genomic DNA, heat-denatured, and then annealed to the same strand of the target DNA fragment. The primer extends and stops until it meets the block probe, and the extension product is then ligated with the block probe by a thermal stable ligase. The above procedure can be repeated n times on the PCR machine using a two-step PCR cycling program. The extension and ligation product is then purified by an exonuclase mixture digestion to remove any DNA fragment with no exonuclease-resistance modification at both ends such as remained primers or probes, primer dimers and genomic DNA, and then amplified using universal NGS PCR primer pairs. A CPELA reaction can include hundreds of extension primer-blocking probe sets and simultaneously amplify hundreds of target DNA fragments for subsequent massively parallel sequencing.
DOI: http://dx.doi.org/10.7554/eLife.06322.010
Figure 3—figure supplement 3. The principle of CNVplex technology.

Figure 3—figure supplement 3.

DOI: http://dx.doi.org/10.7554/eLife.06322.011
Open in a new tab

Large deletion mutations in MVK and FDPS were identified in PK patients

For 27 samples with no confirmed point mutation in MVK, PMVK, MVD, and FDPS genes, large deletion or duplication mutations were interrogated by analyzing the copy number of promoters and all exons in these genes. As a result, three deletion mutations were identified by the CNVplex assay. These three deletions were identified in four (F-26, F-40, S-69 and S-21), one (S-56) and one (F-6) cases, respectively. The breakpoints of three large deletions were also determined (Figure 3, Figure 3—figure supplement 1).

Confirmation of genomic variations

We determined the genomic variations by Sanger sequencing in other family members apart from one with an unknown mutation. In 20 pedigrees with more than two blood samples collected, all patients carried the same mutation with the proband. On the other hand, no mutation was detected in any of the healthy family members. Except for one common MVD mutation of c.746T>C, all deleterious candidate SNVs and large deletions were undetected in healthy family members and 270 unrelated ethnically matched controls. In total, we identified 13 mutations in MVD, six mutations in PMVK, four mutations in FDPS, and 28 mutations in MVK in 113 of the 134 PK patients (Supplementary file 2). Notably, 21 of the 28 MVK mutations are novel (Li and Zhang, 2014). These mutations include 30 missense (58%), two start codon (4%), five nonsense (10%), seven indels (14%), four splicing defects (8%), and three large deletions (6%), and predictably affect functions of the respective enzymes. The two MVD mutations (c.746T>C and c.875A>G) were identified in 50 unrelated patients, accounting for 81% of all patients with MVD mutations. Among the 60 familial patients in whom at least one mutation was found, 30 had MVD mutations, three had PMVK mutations, two had FDPS mutations, 24 had MVK mutations, and one had both MVK and MVD mutations. Among the 53 sporadic patients in whom at least one mutation was found, 31 had MVD mutations, six had PMVK mutations, two had FDPS mutations, and 14 had MVK mutations. No mutation was found in the remaining 21 PK patients.

Solute carrier family 17, member 9 gene (SLC17A9) mutation screening

For the 21 PK patients in whom no pathogenic mutation in MVK, PMVK, MVD, or FDPS was found, we sequenced all exons of the SLC17A9 gene by Sanger sequencing. Mutations in the SLC17A9 gene were recently reported to be another genetic cause for DSAP (Cui et al., 2014). However, no mutation in the SLC17A9 gene was found in the 21 PK patients studied.

Clinical manifestations in the 134 index patients with PK

To explore the genotype–phenotype correlations, we studied the clinical records of 134 index patients (92 males and 42 females) with PK. In all patients, the clinical diagnosis was supported by histopathological examinations. The onset of the disease ranged from birth to 78 years (mean age 31 years). The distribution of lesions could be categorized as localized and disseminated forms (Schamroth et al., 1997). Several classical variants were found to coexist in the same patient. However, no malignant degeneration was found in the lesions of these patients. Although all affected members in the same family carried the same mutation, each affected member showed different clinical manifestations and severity.

Genotype–phenotype correlations in PK patients

We observed some interesting correlations between gene mutations and clinical phenotypes of PK (Table 1 and Figure 4). First, giant plaque-type porokeratosis ptychotropica (PPt) with lesion diameters at least 5 cm appeared to be a unique phenotype associated with MVK mutations. This feature was observed in 50% (19/38) of index patients with MVK mutations, but not in any patients with MVD, PMVK, or FDPS mutations. In addition, patients with MVK mutations generally showed the widest range of phenotypes in terms of both the number and size of lesions. Second, localized genital PK (Chen et al., 2006) and porokeratoma (Walsh et al., 2007) appeared to be unique phenotypes associated with PMVK mutations. Third, in PK patients with MVD mutations, the age of onset spanned from 5 to 70 years, and the diameter of the lesions was generally less than 2 cm. Five females and one male with MVD mutations manifested mild solar facial PK (Sharquie and Al-Baghdady, 2003), which was not found in any patients with mutations in other genes. Fourth, in patients with FDPS mutations, the number of lesions was generally more than 500 and the diameter of the lesions was less than 1 cm.

Table 1.

Clinical characteristics and genetic causes of 134 index patients with PK

DOI: http://dx.doi.org/10.7554/eLife.06322.012

Genetic causes of 134 index patients with PK
MVK (39*) PMVK (9) MVD (62*) FDPS (4) Unknown (21)
Sex
 Male 27 9 38 1 17
 Female 12 0 24 3 4
Number of lesions
 0–10 4 4 0 0 6
 10–100 20 5 13 0 5
 100–500 5 0 39 0 5
 >500† 10 0 10 4 5
Diameter of lesions
 Minimum 2 mm 5 mm 1 mm 1 mm 1 mm
 Maximum 20 cm 5 cm 2 cm 1 cm 2 cm
Age of onset
 At birth 0 0 0 0 3
 0–20 23 4 20 1 0
 20–40 12 3 28 1 9
 40–60 3 1 12 2 5
 >60 1 1 2 0 4
Variants of PK
 DSAP/DSP 26 0 56 4 10
 SFP 0 0 6 0 0
 PM 23 9 0 0 6
 HPM 13 3 1 0 0
 Giant plaque of PPt 19 0 0 0 0
 Genital PK (localized) 0 4 0 0 0
 Porokeratoma 0 5 0 0 0
 LP 1 1 3 0 5
Comorbidity
 Psoriasis vulgaris 4 0 2 0 0
Open in a new tab
*

One PK patient (proband of family-28, female), who has both the mutation c. 235G>A (MVK) and the mutation c. 746T>C (MVD), was included in both MVK and MVD groups.

†

The number of lesions is more than 500.

DSAP, disseminated superficial actinic porokeratosis; DSP, disseminated superficial porokeratosis; HPM, hyperkeratotic porokeratosis, LP, linear porokeratosis, PK, porokeratosis; PM, porokeratosis of Mibelli, PPt, porokeratosis ptychotropica; SFP, solar facial porokeratosis.

Figure 4. Representative clinical phenotypes and histopathology associated with the four genotypes.

Figure 4.

Open in a new tab

From left to right, pedigree charts, clinical phenotypes and the corresponding histopathology photos are shown correspondingly. (A) Family (F)-42 proband with MVK mutation showed giant hyperkeratotic plaque-type porokeratosis ptychotropica. (B) F-60 proband with PMVK mutation showed tumor-like porokeratoma in the genitogluteal region. (C) F-36 proband with MVD mutation showed discrete, red-brown annular keratotic papules or maculopapules on the chest. (D) F-47 proband with FDPS mutation showed multiple, small, superficial, annular papules with thread-like ridges on the legs. All histopathology showed cornoid lamella, a histological hallmark of porokeratosis with vertical columns of parakeratosis overlying an area of hypogranulosis with dyskeratotic cells.

DOI: http://dx.doi.org/10.7554/eLife.06322.013

As a whole, the lesions harboring mutations in MVD and FDPS genes tended to be more homogeneous and superficial than those carrying mutations in MVK and PMVK genes. The phenotypes of MVD and FDPS mutations on the natal cleft were mild and unobvious. However, some patients harboring MVK and PMVK mutations firstly presented with untypical lesions of genitogluteal porokeratosis (GGP).

Reduced expression of wild allele and RNA editing in LTs

The expression of the wild allele in 10 out of 13 LTs was significantly reduced at the 1% significance level (Figures 5–7, Supplementary file 3). In contrast, the expression of the mutant allele was significantly reduced in almost all NNS (except one from F-42) carrying a mutation resulting in nonsense, frameshift, or splicing defect. As for LT from F-31, the mutant to wild allelic ratio in genomic DNA was >3 compared with the ratio of about 1 in the corresponding NNS genomic DNA. However, no deletion or duplication of all MVK exons was detected in F-31, which might indicate that gene conversion of the wild to the mutant allele occurred in the LT (Figure 5). Gene conversion might also occur in the thigh LT of F-43 (Figure 7). No copy number change or CpG island methylation was detected at the corresponding mutated genes in all LTs. A mutation of c.1003G>A (p.Gly335Ser) on the wild allele (T) of MVK: c.1093T>A was identified in the cDNA of the LT without AEI but not the cDNA of NNS collected from the left forearm of F-38 (Figure 6). Surprisingly, this mutation of c.1003G>A was not identified in the genomic DNA of this LT either, which indicated that G-to-A RNA editing occurred on the wild allele of MVK: c.1093T>A. No mutation was observed in cDNA samples from other tissues.

Figure 5. Gene conversion of the wild to mutant allele was identified in a buttock lesion from F-31 carrying a MVK mutation of c.395delT.

Figure 5.

Open in a new tab

(A) The mutant/wild allelic ratios in genomic DNA (gDNA) and complementary DNA (cDNA) of lesional tissue (LT) and neighboring normal-appearing skin (NNS). The quantity of the mutant allele was about threefold and 10-fold more than the quantity of the wild allele in gDNA and cDNA, respectively. (B) The chromatograms of single nucleotide extension targeting the c.395delT mutation using the SNaPshot kit for five DNA samples (blood gDNA, NNS gDNA, LT gDNA, NNS cDNA and LT cDNA). The mutant peak was overpresented in both LT gDNA and LT cDNA. (C) No copy number change in genomic DNA of NNS and LT. PA and PB are probes in the promoter region, E01 to E11 designates exon 1 to exon 11, and A or B indicate two different probes in the same exon. (D) The bisulfite sequencing of NNS gDNA and LT gDNA. No methylation was observed for the targeted CpG sites in the promoter region of MVK.

DOI: http://dx.doi.org/10.7554/eLife.06322.014

Figure 6. G-to-A RNA editing at position 1003 of the wild allele (T) of c.1093T>A in MVK was detected in a left forearm lesion from F-38.

Figure 6.

Open in a new tab

(A) No allelic expression imbalance was observed in lesional tissue (LT) from F-38. (B) A mutation of c.1003G>A was identified in LT cDNA, but not in neighboring normal-appearing skin (NNS) cDNA or LT gDNA. (C) Sequencing the c.1093A and c.1093T allele-specific PCR products indicated the mutant allele(A) of c.1003G>A was in the cis position with the wild allele(T) of c.1093T>A. A-SPCR, c.1093A–specific PCR; T-SPCR, c.1093T-specific PCR.

DOI: http://dx.doi.org/10.7554/eLife.06322.015

Figure 7. Significantly reduced expression of the wild allele in other nine lesion tissues.

Figure 7.

Open in a new tab

The Student t-test was performed to measure the difference in mutant/wild allelic ratios in genomic DNA (gDNA) or complementary DNA (cDNA) from lesion tissue (LT) and neighboring normal-appearing skin (NNS) for each mutation. The test score (p value) is presented above the LT bars. The asterisk (*) designates a significance level of 1%. F-43a and F-43b indicates the tissue sets of the left forearm and left thigh, respectively, from the same F-43 patient.

DOI: http://dx.doi.org/10.7554/eLife.06322.016

Discussion

In this present study we were able to identify mutations in four genes (MVK, PMVK, MVD, and FDPS) underlying PK by massively parallel sequencing and copy number variation analysis in 134 index patients with PK. In the 12 PK families where the affected members span three to six generations, we found mutations only in MVK or MVD but not in PMVK or FDPS (Figure 1, Figure 1—figure supplement 1). We also observed the correlation of genotypes and several clinical manifestations in the PK patients (Table 1), which might be helpful to simplify the classification of PK under the guidance of genetic testing.

Interestingly, two patients were identified to carry two mutations: c.302C>G (p.Pro101Arg) and c.683G>A (p.Arg228Gln) in MVD for S-62, and c.746T>C (p.Phe249Ser) in MVD and c.235G>A (p.Asp79Asn) in MVK for F-28. According to the prediction programs of bioinformatics, the c.235G>A in MVK and c.302C>G in MVD were less conservative than the other mutation carried in F-28 and S-62, respectively. The two mutations in MVD for S-62 were located in the trans position, for his daughter carried only one of them (Figure 1—figure supplement 2). S-62 had late-onset PK at the age of 50 and his 30-year-old daughter had no lesions to date. According to our observation, index patients F-28 and S-62 did not present with more severe phenotypes or earlier age of onset or any other unique clinical features.

In this study we did not detect pathogenic mutation in 21 patients: five cases with linear PK (LP), six cases with solitary plaque-type PK, and 10 cases with disseminated PK (only one had family history). There were three possible explanations: (1) these patients might have other non-classical or mosaic/somatic mutations that could not be detected by current mutation screening methods; (2) these patients might have mutations in unknown genes; (3) these patients might be acquired ISIP without mutation. Recently, mutations in the SLC17A9 gene were reported to be another genetic cause for DSAP (Cui et al., 2014). We sequenced all exons of SLC17A9 in these 21 patients by Sanger sequencing but found no mutations in SLC17A9.

RNA editing is a process of some discrete changes to specific nucleotide sequences of a RNA molecule which may be critical during normal development and diseases (Witkin, et al., 2015). G-to-A RNA editing in PK was first reported in our study. Interestingly, the RNA editing seemed to occur selectively on the one copy of MVK and this selective editing pattern seemed to be capable of being transmitted during mitosis. The mechanism underlying this observation should be investigated further.

Using AEI assays, we observed that the expression of wild allele was significantly reduced to various extents in most LTs (>77%). The variation in the expression reduction level might be due to the LT containing various amounts of normal-appearing keratinocytes and other different types of cells. Somatic loss of function of the wild allele in the epidermal stem cells might play a key role in the formation of lesions, and it might occur spontaneously and be induced by environmental factors such as ultraviolet light.

There was only one RNA editing identified in the LT, therefore the somatic loss of function of the wild allele in LTs of PK could be mainly caused by suppression of the wild allele through gene conversion in a minor proportion of cases or an unknown DNA methylation-independent epigenetic mechanism in most cases. Allelic imbalance of histone methylation modification unlinked to DNA methylation, whose pattern is able to be transmitted during cell mitosis, could be the DNA methylation-independent epigenetic mechanism (Xu et al., 2011). Further investigation would be necessary to test this hypothesis. We did not find reduced expression or somatic mutation of the wild allele in only two LTs. The reason for this could be that the LT might contain many normal cells affecting the sensitivity of AEI and mutation detection.

As a starter button of PK, gene conversion of the wild allele to the mutant allele was first identified in this study. This observation is generally consistent with the phenomenon known as ‘revertant mosaicism’, in which the mutant allele is corrected in some stem cells. Revertant mosaicism is mainly identified in blood diseases such as Wiskott–Aldrich syndrome and Fanconi’s anemia, and skin diseases such as ichthyosis and epidermolysis bullosa (May, 2011). Revertant mosaicism shows the bright side of gene conversion, but our observation revealed the opposite side. Therefore, gene conversion could be a natural ‘rescue angel’ in some recessive diseases or a ‘damaging devil’ in some dominant diseases, which could mostly affect tissues with extensive ability to self-renew for life.

Our findings support the notion that Mendelian diseases can be caused by multiple genes involved in the same metabolic or signal transduction pathway, or producing proteins on interaction. The four genes MVK, PMVK, MVD, and FDPS are members of the mevalonate pathway, suggesting that this pathway is involved in the pathogenesis of PK. The mevalonate pathway of isoprenoid biosynthesis provides precursors of isoprenoids, which are ubiquitous in living species and diverse in biological function (Smit and Mushegian, 2000) and serve as precursors of cholesterol, heme A, ubiquinones, dolichol, and isoprenylated proteins (e.g., RAS and Lamin B) (Figure 2). Moreover, it is also the subject of many pharmacological regulating drugs such as statins. It is well known that cholesterol is an important constituent of the cell membrane of most eukaryotic cells, and several inherited disorders have been linked to defects in cholesterol biosynthesis. The isoprenylated proteins play a role in the regulation of cell growth, division, and differentiation (Moir and Spann, 2001; Zwerger et al., 2011; Davidson and Lammerding, 2014) and are probably associated with the retained nuclei in the stratum corneum (i.e., parakeratosis). We propose that the accumulation of abnormal metabolites or shortage of isoprenoids might predispose patients to idiopathic inflammation of the skin, and that correction of the abnormal isoprenoid biosynthesis might be a novel therapeutic direction for PK. Further functional studies will be necessary to confirm this hypothesis.

Materials and methods

Ethics statement

All procedures followed the guidelines of the Helsinki Declaration and were approved by the ethics committee and by the Scientific Ethical Committee of Fudan University. Participating centers provided local Institutional Review Board approval for genetic analysis. Study participants provided informed consent for genetic testing.

Clinical samples

From 2001 to 2013, peripheral blood samples were collected from 134 index patients with PK (92 males and 42 females) and 180 family members (103 males and 77 females). Thirteen pairwise LTs and NNS were dissected from 12 index patients with a MVK, MVD, or FDPS mutation (Supplementary file 3). In addition, blood samples were collected from 270 healthy adult individuals undergoing routine medical examination in the hospital. The 134 index patients were diagnosed by at least two experienced dermatologists, based on both clinical features and histological examinations. In view of the family history, these index patients were divided into family (F) and sporadic (S) cases. Among the 61 familial cases, 40 index patients were probands from each family. We examined members of 21 families from different provinces of China, whose generations ranged from 2 to 6.

DNA and RNA preparation

Genomic DNA was extracted from the peripheral blood using the QIAamp DNA blood mini kit (Qiagen, Germany). Total RNA was isolated from the skin using the RNeasy Protect Mini Kit (Qiagen, Germany) according to the manufacturer's instructions.

Sanger sequencing and SNV genotyping

PCR products were sequenced by BigDye3.1 (Applied Biosystems, Foster City, CA, USA). Raw data were collected on an ABI 3130XL sequencer and mutations were called by the PolyPhred program (Nickerson et al., 1997). The 60 SNVs in 270 control samples were genotyped using the ABI PRISM SNaPshot Multiplex Kit (Applied Biosystems, Foster City, CA, USA) and ABI3730xl sequencer.

CPELA

A total of 162 fragments covering the promoter region, 5′UTR, coding sequence and splicing site of 12 genes in the mevalonate pathway (ACAT1, ACAT2, HMGCS1, HMGCS2, HMGCR, MVK, PMVK, MVD, IDI1, IDI2, FDPS, and GGPS1) were amplified using the EasyTarget® amplification kit (Genesky Biotechnologies Inc, Shanghai, China) which was developed according to the CPELA method. The CPELA was a new method developed by Genesky Biotechnologies for fast and simple enrichment of multiple gene regions for massively parallel sequencing. The principle of CPELA is described in Figure 3—figure supplement 2.

Massively parallel sequencing

The amplification products by CEPLA were mixed and size-separated by 2% agarose gel electrophoresis. Products of length 200–400bp were recovered. The final concentration of the library mixture was determined by real-time quantitative PCR and the average concentration for each library was estimated. The libraries tagged with different index sequences from several projects were well mixed in an appropriate concentration for each library corresponding to its aimed sequencing depth. The final library mixture was sequenced on MiSeq sequencer (Illumina) using Miseq reagent kit v2. The sequencing reads were separated for each sample by running CASAVA (Illumina Inc, San Diego, CA, USA) and target reads were determined by comparing them with fragment reference sequences (hg19) using the Blat program (Kent, 2002).

Bioinformatics

In order to reduce the reading error from the sequencing reaction step, we first compared the two reads from the same cluster and integrated them into one correction read. The correction reads were then aligned to hg19 using the Burrows–Wheeler Aligner (BWA) (Li and Durbin, 2010). SNV calling was performed using both GATK (McKenna et al., 2010) and Varscan programs (Koboldt et al., 2012), and the called SNV data were then combined. The Annovar program was used for SNV annotation (Wang et al., 2010). The functional effect of non-synonymous SNVs was assessed by the PolyPhen-2, SIFT, and MutationTaster (Ng and Henikoff, 2003; Adzhubei et al., 2010; Schwarz et al., 2010). Non-synonymous SNVs with SIFT score of <0.05, Polyphen-2 score of >0.85 or MutationTaster score of >0.85 were considered as significant of not being benign. To sort potentially deleterious variants from benign polymorphisms, perl scripts were used to filter the SNVs against those of dbSNP135. Any SNV recorded in dbSNP135 and with a minor allele frequency of ≥1% in Chinese from 1000 genome database was considered as benign polymorphisms and therefore removed for subsequent analysis.

CNVplex assay and breakpoint analysis

The copy number of the target regions was measured by a CNVplex assay, a high-throughput multiplex CNV analysis method recently developed by Genesky Biotechnologies. The principle of CNVplex technology is described in Figure 3—figure supplement 3. We utilized this technology for quantitative analysis of copy numbers of all 37 exons and upstream promoter regions in the MVK, MVD, PMVK, and FDPS genes for the blood DNA samples with no point mutation identified and the tissue DNA samples. Based on the copy number measurements for all target sequences, the breakpoints were estimated to be located between two neighboring probe target sites showing different copy numbers. Several primer sets flanking the two probe target sites were tested to amplify the target region from case and control DNA samples using a long PCR protocol. Specific PCR products from case samples were sequenced using the ABI BigDye3.1 and the breakpoints were determined by blasting the sequences with human reference genome assembly.

AEI assay

Both DNA and RNA were extracted from pairwise LT and NNS. Each RNA sample was reversely transcripted into cDNA twice using polyA and N9 primer mix and Reverse Transcriptase M-MLV (RNase H-) (New England Biolabs, England). For each pairwise tissue set, single nucleotide extension was used to quantitate the ratio of the mutant to the wild allele both in the tissue cDNA and DNA using the ABI PRISM SNaPshot Multiplex Kit (Applied Biosystems), followed by normalization to the ratio value in the corresponding patient’s blood DNA as a reference of 1:1. As for cDNA from F-42 and S-36 patients with splicing defect mutations of MVK: c.371+2T>A and FDPS: c.684+1G>A respectively, fluorescent PCR followed by capillary electrophoresis was used to quantitate the ratios of mutant to wild transcripts.

cDNA sequencing and bisulfite sequencing

Total RNA was reversely transcribed into cDNA. Two or three sets of primers were designed to amplify the cDNA of MVK, MVD, or FDPS, covering all coding sequences for each gene. The sequences of PCR products were determined by Sanger sequencing. Two allele-specific PCRs for c.1093T>A followed by Sanger sequencing were performed in order to determine whether the mutant allele (A) of c.1003G>A was in the cis position with the wild allele (T) of c.1093T>A. DNA from each pairwise LT and NNS were subject to bisulfite conversion using the EZ DNA Methylation Kit (Zymo Research Corporation, Irvine, CA, USA). PCR products amplifying the bisulfite converted DNA at CpG islands of MVK, MVD, or FDPS were sequenced on MiSeq sequencer (Illumina) using Miseq reagent kit v2 or on the ABI 3130xl genetic analyzer using the ABI BigDye3.1.

Acknowledgements

We thank all patients and family members who participated in this study. This work is sponsored by grants from the Shanghai Pujiang Program, the Research Special Fund for Public Welfare Industry of Health, Guideline-oriented Research in the Management of Some Common and Severe Skin Diseases (No. 201002016), Construction of National Key Clinical Specialty in Shanghai, Ministry of Health. Special thanks should be given to Dr Lihua Zheng for helpful discussion.

Funding Statement

The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.

Funding Information

This paper was supported by the following grants:

  • Shanghai Pujiang Program none to Zhenghua Zhang.

  • Research special fund for public welfare of health 201002016 to Zhenghua Zhang.

  • Ministry of Health of the People's Republic of China Construct Construction of National Key Clinical Specialty in Shanghai, Ministry of Health to Zhenghua Zhang.

Additional information

Competing interests

The authors declare that no competing interests exist.

Author contributions

ZZ, Conception and design, Collection of patients, Acquisition of data, Analysis and interpretation of data, Drafting and revising the article.

CL, Conception and design, Acquisition of data, Analysis and interpretation of data, Drafting or revising the article.

ZJ, Conception and design, Acquisition of data, Analysis and interpretation of data, Drafting or revising the article.

FW, Acquisition of data.

JL, Acquisition of data.

WL, Acquisition of data.

SZ, Acquisition of data.

JG, Acquisition of data.

MF, Acquisition of data.

XM, Acquisition of data.

NS, Acquisition of data.

XB, Acquisition of data.

CG, Acquisition of data.

ZZ, Acquisition of data.

QH, Acquisition of data.

LC, Acquisition of data.

LX, Acquisition of data.

JX, Acquisition of data.

ZZ, Acquisition of data.

RM, Acquisition of data, Analysis and interpretation of data.

FY, Acquisition of data, Analysis and interpretation of data.

LW, Acquisition of data, Analysis and interpretation of data.

YL, Acquisition of data, Analysis and interpretation of data.

WH, Acquisition of data, Analysis and interpretation of data.

HP, Acquisition of data, Analysis and interpretation of data.

Ethics

Human subjects: All procedures followed the guidelines of the Helsinki Declaration and were approved by the ethics committee and by the Scientific Ethical Committee of Fudan University. Participating centers provided local Institutional Review Board approval for genetic analysis. Study participants provided informed consent for genetic testing.

Additional files

Supplementary file 1.

12 non-pathogenic rare missense or nonsense single nucleotide variants (SNVs) and major reasons for exclusion.

DOI: http://dx.doi.org/10.7554/eLife.06322.017

elife06322s002.docx (28.7KB, docx)
DOI: 10.7554/eLife.06322.017
Supplementary file 2.

Characterization of 51 mutations identified in 113 of the 134 porokeratosis (PK) patients.

DOI: http://dx.doi.org/10.7554/eLife.06322.018

elife06322s003.docx (42.4KB, docx)
DOI: 10.7554/eLife.06322.018
Supplementary file 3.

Mutant to wild allelic ratio measurements in both genomic DNA and complementary DNA of 13 pairwise tissue sets from 12 patients.

DOI: http://dx.doi.org/10.7554/eLife.06322.019

elife06322s004.docx (47.6KB, docx)
DOI: 10.7554/eLife.06322.019

References

  1. Abramovits W, Oquendo M. Introduction to autoinflammatory syndromes and diseases. Dermatologic Clinics. 2013;31:363–385. doi: 10.1016/j.det.2013.04.010. [DOI] [PubMed] [Google Scholar]
  2. Adzhubei IA, Schmidt S, Peshkin L, Ramensky VE, Gerasimova A, Bork P, Kondrashov AS, Sunyaev SR. A method and server for predicting damaging missense mutations. Nature Methods. 2010;7:248–249. doi: 10.1038/nmeth0410-248. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Biswas A. Cornoid lamellation revisited: apropos of porokeratosis with emphasis on unusual clinicopathological variants. The American Journal of Dermatopathology. 2015;37:145–155. doi: 10.1097/DAD.0000000000000039. [DOI] [PubMed] [Google Scholar]
  4. Chen TJ, Chou YC, Chen CH, Kuo TT, Hong HS. Genital porokeratosis: a series of 10 patients and review of the literature. The British Journal of Dermatology. 2006;155:325–329. doi: 10.1111/j.1365-2133.2006.07295.x. [DOI] [PubMed] [Google Scholar]
  5. Cui H, Li L, Wang W, Shen J, Yue Z, Zheng X, Zuo X, Liang B, Gao M, Fan X, Yin X, Shen C, Yang C, Zhang C, Zhang X, Sheng Y, Gao J, Zhu Z, Lin D, Zhang A, Wang Z, Liu S, Sun L, Yang S, Cui Y, Zhang X. Exome sequencing identifies SLC17A9 pathogenic gene in two Chinese pedigrees with disseminated superficial actinic porokeratosis. Journal of Medical Genetics. 2014;51:699–704. doi: 10.1136/jmedgenet-2014-102486. [DOI] [PubMed] [Google Scholar]
  6. Davidson PM, Lammerding J. Broken nuclei—lamins, nuclear mechanics, and disease. Trends in Cell Biology. 2014;24:247–256. doi: 10.1016/j.tcb.2013.11.004. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Kanehisa M, Goto S, Sato Y, Furumichi M, Tanabe M. KEGG for integration and interpretation of large-scale molecular data sets. Nucleic Acids Research. 2012;40:D109–D114. doi: 10.1093/nar/gkr988. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Kent WJ. BLAT—the BLAST-like alignment tool. Genome Research. 2002;12:656–664. doi: 10.1101/gr.229202. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Koboldt DC, Zhang Q, Larson DE, Shen D, McLellan MD, Lin L, Miller CA, Mardis ER, Ding L, Wilson RK. VarScan 2: somatic mutation and copy number alteration discovery in cancer by exome sequencing. Genome Research. 2012;22:568–576. doi: 10.1101/gr.129684.111. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Li H, Durbin R. Fast and accurate long-read alignment with Burrows-Wheeler transform. Bioinformatics. 2010;26:589–595. doi: 10.1093/bioinformatics/btp698. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Li C, Zhang Z. Updating and correcting the description of MVK mutations identified in Chinese patients with disseminated superficial actinic porokeratosis. Journal of Dermatological Science. 2014;75:68–69. doi: 10.1016/j.jdermsci.2014.03.006. [DOI] [PubMed] [Google Scholar]
  12. Luan J, Niu Z, Zhang J, Crosby ME, Zhang Z, Chu X, Wang Z, Huang W, Xiang L, Zheng Z. A novel locus for disseminated superficial actinic porokeratosis maps to chromosome 16q24.1-24.3. Human Genetics. 2011;129:329–334. doi: 10.1007/s00439-010-0929-x. [DOI] [PubMed] [Google Scholar]
  13. May M. Mutations to the rescue. Nature Methods. 2011;17:405–407. doi: 10.1038/nm0411-405. [DOI] [PubMed] [Google Scholar]
  14. McKenna A, Hanna M, Banks E, Sivachenko A, Cibulskis K, Kernytsky A, Garimella K, Altshuler D, Gabriel S, Daly M, DePristo MA. The genome analysis toolkit: a MapReduce framework for analyzing next-generation DNA sequencing data. Genome Research. 2010;20:1297–1303. doi: 10.1101/gr.107524.110. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Moir RD, Spann TP. The structure and function of nuclear lamins: implications for disease. Cellular and Molecular Life Sciences. 2001;58:1748–1757. doi: 10.1007/PL00000814. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Ng PC, Henikoff S. SIFT: predicting amino acid changes that affect protein function. Nucleic Acids Research. 2003;31:3812–3814. doi: 10.1093/nar/gkg509. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Nickerson DA, Tobe VO, Taylor SL. PolyPhred: automating the detection and genotyping of single nucleotide substitutions using fluorescence-based resequencing. Nucleic Acids Research. 1997;25:2745–2751. doi: 10.1093/nar/25.14.2745. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Schamroth JM, Zlotogorski A, Gilead L. Porokeratosis of Mibelli. Overview and review of the literature. Acta Dermato-Venereologica. 1997;77:207–213. doi: 10.2340/0001555577207213. [DOI] [PubMed] [Google Scholar]
  19. Schwarz JM, Rodelsperger C, Schuelke M, Seelow D. MutationTaster evaluates disease-causing potential of sequence alterations. Nature Methods. 2010;7:575–576. doi: 10.1038/nmeth0810-575. [DOI] [PubMed] [Google Scholar]
  20. Sertznig P, von Felbert V, Megahed M. Porokeratosis: present concepts. Journal of the European Academy of Dermatology and Venereology. 2012;26:404–412. doi: 10.1111/j.1468-3083.2011.04275.x. [DOI] [PubMed] [Google Scholar]
  21. Sharquie KE, Al-Baghdady BA. Solar facial porokeratosis. The Journal of Dermatology. 2003;30:216–221. doi: 10.1111/j.1346-8138.2003.tb00374.x. [DOI] [PubMed] [Google Scholar]
  22. Smit A, Mushegian A. Biosynthesis of isoprenoids via mevalonate in Archaea: the lost pathway. Genome Research. 2000;10:1468–1484. doi: 10.1101/gr.145600. [DOI] [PubMed] [Google Scholar]
  23. Thurnher M, Gruenbacher G, Nussbaumer O. Regulation of mevalonate metabolism in cancer and immune cells. Biochimica et Biophysica Acta. 2013;1831:1009–1015. doi: 10.1016/j.bbalip.2013.03.003. [DOI] [PubMed] [Google Scholar]
  24. Walsh SN, Hurt MA, Santa Cruz DJ. Porokeratoma. The American Journal of Surgical Pathology. 2007;31:1897–1901. doi: 10.1097/PAS.0b013e31806910c7. [DOI] [PubMed] [Google Scholar]
  25. Wang K, Li M, Hakonarson H. ANNOVAR: functional annotation of genetic variants from high-throughput sequencing data. Nucleic Acids Research. 2010;38:e164. doi: 10.1093/nar/gkq603. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Witkin KL, Hanlon SE, Strasburger JA, Coffin JM, Jaffrey SR, Howcroft TK, Dedon PC, Steitz JA, Daschner PJ, Read-Connole E. RNA editing, epitranscriptomics, and processing in cancer progression. Cancer Biology & Therapy. 2015;16:21–27. doi: 10.4161/15384047.2014.987555. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Xu M, Wang W, Chen S, Zhu B. A model for mitotic inheritance of histone lysine methylation. EMBO Reports. 2011;13:60–67. doi: 10.1038/embor.2011.206. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Zeng K, Zhang QG, Li L, Duan Y, Liang YH. Splicing mutation in MVK is a cause of porokeratosis of Mibelli. Archives of Dermatological Research. 2014;306:749–755. doi: 10.1007/s00403-014-1465-7. [DOI] [PubMed] [Google Scholar]
  29. Zhang SQ, Jiang T, Li M, Zhang X, Ren YQ, Wei SC, Sun LD, Cheng H, Li Y, Yin XY, Hu ZM, Wang ZY, Liu Y, Guo BR, Tang HY, Tang XF, Ding YT, Wang JB, Li P, Wu BY, Wang W, Yuan XF, Hou JS, Ha WW, Wang WJ, Zhai YJ, Wang J, Qian FF, Zhou FS, Chen G, Zuo XB, Zheng XD, Sheng YJ, Gao JP, Liang B, Li P, Zhu J, Xiao FL, Wang PG, Cui Y, Li H, Liu SX, Gao M, Fan X, Shen SK, Zeng M, Sun GQ, Xu Y, Hu JC, He TT, Li YR, Yang HM, Wang J, Yu ZY, Zhang HF, Hu X, Yang K, Wang J, Zhao SX, Zhou YW, Liu JJ, Du WD, Zhang L, Xia K, Yang S, Wang J, Zhang XJ. Exome sequencing identifies MVK mutations in disseminated superficial actinic porokeratosis. Nature Genetics. 2012;44:1156–1160. doi: 10.1038/ng.2409. [DOI] [PubMed] [Google Scholar]
  30. Zhang ZH, Wang ZM, Crosby ME, Wang HF, Xiang LH, Luan J, Gu CY, Zhou J, Niu ZM, Fang X, Huang W, Zheng ZZ. Gene expression profiling of porokeratosis. Journal of Cutaneous Pathology. 2008;35:1058–1062. doi: 10.1111/j.1600-0560.2008.01026.x. [DOI] [PubMed] [Google Scholar]
  31. Zwerger M, Ho CY, Lammerding J. Nuclear mechanics in disease. Annual Review of Biomedical Engineering. 2011;13:397–428. doi: 10.1146/annurev-bioeng-071910-124736. [DOI] [PMC free article] [PubMed] [Google Scholar]
eLife. 2015 Jul 23;4:e06322. doi: 10.7554/eLife.06322.020

Decision letter

Editor: Helen H Hobbs1

eLife posts the editorial decision letter and author response on a selection of the published articles (subject to the approval of the authors). An edited version of the letter sent to the authors after peer review is shown, indicating the substantive concerns or comments; minor concerns are not usually shown. Reviewers have the opportunity to discuss the decision before the letter is sent (see review process). Similarly, the author response typically shows only responses to the major concerns raised by the reviewers.

Thank you for sending your work entitled “Genomic Variations of the Mevalonate Pathway in Porokeratosis” for consideration at eLife. Your article has been favorably evaluated by Stylianos Antonarakis (Senior editor) and three reviewers, one of whom, Helen Hobbs, is a member of our Board of Reviewing Editors.

The Reviewing editor and the other reviewers discussed their comments before we reached this decision, and the Reviewing editor has assembled the following comments to help you prepare a revised submission.

This paper explores the genetic basis of porokeratosis, a genetically heterogeneous autosomal dominant disorder of keratinization that includes both inherited and sporadic forms. Mutations in mevalonate kinase (MVK), a key enzyme in sterol synthesis, were shown previously by these authors to cause the disease in a subset of individuals, but the etiology in most affected individuals has remained unknown. Here the authors test the hypothesis that the disease is due to defective isoprenoid generation and function. First, they used linkage analysis followed by candidate gene sequencing to identify a second defective gene in the mevalonate pathway that causes this disorder, mevalonate decarboxylase (MVD). They then sequenced 12 genes in the mevalonate and isoprenoid pathways in 134 index cases and provide evidence that mutations inactivating 3 more genes in the pathway (PMVK, MVD and FDPS) also cause this disease. Collectively, implicated variants in the 4 genes identified explain disease in 98% of familial cases and 73% of apparently sporadic cases. They also provide evidence for correlation between the underlying gene and selected aspects of clinical presentation.

The three reviewers agree that the paper is clearly written and the major conclusions of the paper are supported by the data presented. Although this represents an important advance in the molecular characterization of this disorder, and further implicates and refines the disrupted pathway, the paper does not provide sufficient mechanistic insights to warrant publication in eLife and without such data is better suited for a subspecialty journal. How do defects in mevalonate metabolism cause the keratinization defects – toxic intermediary metabolites vs. deficiency of metabolic end-products? Most individuals were heterozygous for a mutation. Since the genes all encode enzymes, is there a mutation in the normal allele, or possibly in another allele in the pathway, in the skin lesions? Providing data to address any one of these questions would suffice.

Other major concerns:

1) While the authors describe that complete co-segregation between each putative mutation and disease was observed (and representative pedigrees are shown), these data should be expanded. While it might not be plausible to show pedigrees for all 60 familial cases with a defined mutation, it would be worth noting the size distribution of families for each gene and the number of families that achieved an independent LOD score >2.0 or >3.0, for example.

2) Curiously, it is stated that the vast majority of cases associated with variation in PMVK or FDPS were sporadic. If familial segregation was ever seen for these 2 genes, it should be shown. Genetic evidence that at least some of the sporadic mutations are de novo should be provided. It would also be informative to perform a formal burden test for these 2 genes to document enrichment for novel/rare and predicted deleterious variation in association with phenotype.

3) There is no evidence of pathogenicity for a number of the missense variants listed in Table 1 aside from the fact that they are present in cases and not controls (or 1 control, in the case of variant 44, MVD c.746T>C in Table 1). Providing information regarding the presence of these mutations in the public databases would be informative. A formal burden test for each new gene should be performed.

4) While the authors acknowledge that MVK mutations have previously been observed in other disorders, including mevalonic aciduria (MA) and hyper IgD syndrome (HIGDS), the same appears not to be true for the other 3 genes. Curiously, patients with biallelic mutations causing MA or HIGDS do not show porokeratosis, nor do their heterozygous parents. This is particularly puzzling because this group has identified porokeratosis in individuals heterozygous for some of the same mutations. Variation in environmental exposure (e.g. UV light) has been invoked as a potential mechanism, however this seems less tenable given the complete penetrance seen for porokeratosis in the described families. Somatic loss of the other MVK allele in skin lesions was previously excluded in a handful of patients, but this should be revisited in this manuscript.

eLife. 2015 Jul 23;4:e06322. doi: 10.7554/eLife.06322.021

Author response

1) While the authors describe that complete co-segregation between each putative mutation and disease was observed (and representative pedigrees are shown), these data should be expanded. While it might not be plausible to show pedigrees for all 60 familial cases with a defined mutation, it would be worth noting the size distribution of families for each gene and the number of families that achieved an independent LOD score >2.0 or >3.0, for example.

Among 61 familial cases, we examined members of 21 families from different provinces of China, whose generations ranged from 2 to 6. There were 5 core families (F-1, F-2, F-3, F-19 and F-24) in which more than 10 family members and blood samples were collected. The maximum two-point LOD score of F-1, F-2, F-3, F-19 and F-24 were 5.56 (D12S79), 8.62 (D12S84), 3.73 (D16S3074), 2.78 (D12S1583), 4.61 (D12S84), respectively. There were 40 probands from different PK families during the past 10 years, which could have been used for confirmation. All of them were adult and came to the out-patient department by themselves. We didn’t examine their family members and collected their blood samples. Therefore, it is difficult to calculate a reasonable LOD score in those non-core pedigrees. The sentence “We confirmed the genomic variations by Sanger sequencing in each family member. Each mutation displayed 100% cosegregation with PK phenotype in the family” may be misleading. Therefore, it was replaced with: “We determined the genomic variations by Sanger sequencing in other family members apart from one with unknown mutation. In 20 pedigrees with more than two blood samples collected, all patients carried the same mutation with the proband, on the contrary, no mutation was detected in all family healthy members”.

2) Curiously, it is stated that the vast majority of cases associated with variation in PMVK or FDPS were sporadic. If familial segregation was ever seen for these 2 genes, it should be shown. Genetic evidence that at least some of the sporadic mutations are de novo should be provided. It would also be informative to perform a formal burden test for these 2 genes to document enrichment for novel/rare and predicted deleterious variation in association with phenotype.

The sentence “In contrast, mutations in PMVK and FDPS were mainly found in proband or sporadic cases” was misleading. Since only 9 index patients were identified to carry PMVK mutations, and 3 of them were familial cases, and only 4 index patients were identified to carry FDPS mutations, and 2 of them were familial cases. Those index patients with negative family history were classified as the sporadic cases. Actually, the sporadic case might not be truly the only one patient in his family, because PK could be late-onset, undiscovered in non-exposure region. Therefore, we deleted this sentence in our revised manuscript. All 73 sporadic cases were adult and collected during the past 10 years, which could have been used for confirmation. Most of them lived separately with their parents and were out of touch with us for years. We didn’t examine those parents and get no informed consent and no peripheral blood from them. We are really sorry for that. The formal burden test is generally used for complex diseases with a large number of samples (Nat Genet. 2012, 29;44:623-630; Hum Mol Genet. 2012,21:R1-9). However, it is not suitable for our study.

3) There is no evidence of pathogenicity for a number of the missense variants listed in Table 1 aside from the fact that they are present in cases and not controls (or 1 control, in the case of variant 44, MVD c.746T>C in Table 1). Providing information regarding the presence of these mutations in the public databases would be informative. A formal burden test for each new gene should be performed.

The highest frequency mutation (c.746T>C in MVD) was identified in one control. This control was a 22 year-old male, who took routine medical examination in the hospital. He had no PK lesions till now and gave negative familial history of skin diseases. He could be predicted to develop PK in future. In view of its high frequency of occurence in PK, p-value of this mutation showed significant difference (p=0.00). Including this common mutation, there were 11 mutations had been identified in at least 2 index patients, but not in 270 controls. There are 7 mutations in MVK have been reported as indicated in Table 1, moreover, 5 mutations of them were identified in at least 2 index patients and 2 mutations of them were identified in 2 different sporadic cases. There were 4 mutations identified only in one family, but more than 2 patients in the pedigree were identified to carry this mutation and other healthy family members were not. Thus, at least 17 mutations in MVK, MVD and PMVK could be considered as pathogenic with strong evidence.

A formal burden test for each new gene may not be suitable for our study for the following two reasons: 1) the genetic and phenotypic information could provide solid support of MVK, MVD and PMVK being pathogenic genes of PK, especially MVK and MVD; 2) A formal burden test is generally used for complex traits with a large number of samples (Nat Genet. 2012, 29;44:623-630; Hum Mol Genet. 2012,21:R1-9). However, PK is regarded as monogenic disease with obvious dominant inheritance and a pedigree-based analysis could provide a more powerful way to identify the pathogenic genes, moreover, there were only 9 and 4 index patients with PK carrying different mutations in PMVK and FDPS, respectively.

4) While the authors acknowledge that MVK mutations have previously been observed in other disorders, including mevalonic aciduria (MA) and hyper IgD syndrome (HIGDS), the same appears not to be true for the other 3 genes. Curiously, patients with biallelic mutations causing MA or HIGDS do not show porokeratosis, nor do their heterozygous parents. This is particularly puzzling because this group has identified porokeratosis in individuals heterozygous for some of the same mutations. Variation in environmental exposure (e.g. UV light) has been invoked as a potential mechanism, however this seems less tenable given the complete penetrance seen for porokeratosis in the described families. Somatic loss of the other MVK allele in skin lesions was previously excluded in a handful of patients, but this should be revisited in this manuscript.

Mevalonate kinase deficiency syndrome (MKD) are early onset (usually <12 months), and it was reported that most hyper IgD syndrome (HIGDS) showed mild features of vasculitis (Arch Dermatol. 1994; 130: 59-65). Actually, it is a puzzle that why the patients with MA and HIGDS do not show PK. From the database of hereditary auto-inflammatory disorders mutations, c.417dupC and c.604G>A in MVK were both reported in HIDS and PK. As for those heterozygous parents, the possible reason was that pediatricians and rheumatologists might neglect to examine their skin, since they were not familiar with porokeratosis.

Somatic deletion of the wild MVK allele in skin lesions was not detected in our samples either, however, we observed the reduced expression of the wild allele in most skin lesions and a G-to-A RNA editing on the wild allele in one skin lesion. Therefore, the somatic loss of function, but not somatic deletion, of the wild allele in skin lesions should be common.

Associated Data

    This section collects any data citations, data availability statements, or supplementary materials included in this article.

    Supplementary Materials

    Figure 3—source data 1. Sanger sequencing chromatograms of normal control and porokeratosis (PK) patients at 48 mutation sites in MVK, PMVK, MVD and FDPS.

    DOI: http://dx.doi.org/10.7554/eLife.06322.008

    elife06322s001.pdf (966.6KB, pdf)
    DOI: 10.7554/eLife.06322.008
    Supplementary file 1.

    12 non-pathogenic rare missense or nonsense single nucleotide variants (SNVs) and major reasons for exclusion.

    DOI: http://dx.doi.org/10.7554/eLife.06322.017

    elife06322s002.docx (28.7KB, docx)
    DOI: 10.7554/eLife.06322.017
    Supplementary file 2.

    Characterization of 51 mutations identified in 113 of the 134 porokeratosis (PK) patients.

    DOI: http://dx.doi.org/10.7554/eLife.06322.018

    elife06322s003.docx (42.4KB, docx)
    DOI: 10.7554/eLife.06322.018
    Supplementary file 3.

    Mutant to wild allelic ratio measurements in both genomic DNA and complementary DNA of 13 pairwise tissue sets from 12 patients.

    DOI: http://dx.doi.org/10.7554/eLife.06322.019

    elife06322s004.docx (47.6KB, docx)
    DOI: 10.7554/eLife.06322.019

    Articles from eLife are provided here courtesy of eLife Sciences Publications, Ltd

    RESOURCES