| Human Molecular Genetics | Pages |
©1999 Oxford University Press |
Spectrum of novel ATP2A2 mutations in patients with Darier's disease
Introduction
Results
Spectrum of mutations
Haplotype analysis
Genotype-phenotype comparison
Discussion
Predicted effects of ATP2A2 mutations on SERCA2 function
Distribution of ATP2A2 mutations
Genotype-phenotype correlations
Materials And Methods
Patients
PCR and RT-PCR amplification
CSGE analysis
Sequence analysis
Haplotype analysis
Acknowledgements
References
Spectrum of novel ATP2A2 mutations in patients with Darier's disease
Received May 17, 1999; Revised and Accepted June 24, 1999
Darier's disease (DD) is an autosomal dominantly inherited skin disorder characterized by loss of adhesion between epidermal cells (acantholysis) and abnormal keratinization. Recently, we identified ATP2A2 encoding the sarco/endoplasmic reticulum Ca2+ ATPase isoform 2 as the defective gene in DD. Now we report a spectrum of ATP2A2 mutations in 19 families and six sporadic cases with DD and investigate genotype-phenotype correlations. All 21 exons and flanking intron boundaries were amplified and screened for mutations by conformation-sensitive gel electrophoresis and direct sequencing. We identified 24 novel mutations that are scattered throughout the ATP2A2 gene. Two families shared an identical mutation on a common disease-associated haplotype, suggesting inheritance from a common ancestor. The majority of the mutations (54%; 13/24) led to a premature termination codon which further supports the proposal that haploinsufficiency is a common molecular mechanism for DD. Thirty-eight per cent of mutations (9/24) result in non-conservative amino acid substitutions at highly conserved positions. Two mutations predict mutated polypeptides lacking or carrying additional amino acids. Marked inter- and intrafamilial phenotypic variability of the disease was observed. These results illustrate the considerable diversity of ATP2A2 mutations causing DD and suggest that additional factors are important contributors to the clinical phenotype.
INTRODUCTION
Darier's disease (DD; OMIM 124200) is an autosomal dominant skin disorder characterized by loss of adhesion between epidermal cells (acantholysis) and abnormal keratinization (1). Prevalence has been estimated at 1 in 55 000. DD patients develop warty papules and plaques on the central trunk, scalp, forehead and flexures (Fig. 1a). Most patients also have pits or keratotic papules on the palms and distinctive nail abnormalities; some patients have oral lesions. Heat, sweating and sunburn exacerbate disease symptoms. Onset is usually around puberty and penetrance is complete but the phenotype is variable (2). Patients with mild disease may have no more than a few scattered keratotic papules or subtle nail changes, whereas those with severe disease are handicapped by widespread malodorous keratotic plaques. Some patients have blisters and painful erosions. Oral retinoids reduce hyperkeratosis and malodour in most patients. In a few families, neuropsychiatric abnormalities such as mild mental retardation and epilepsy have been reported (1). The typical histological features include focal areas of separation between suprabasal epidermal cells and abnormal keratinization. Immunohistopathology and electron microscopy reveal loss of desmosomal attachments and perinuclear aggregations of keratin filaments (3). These observations suggested that defects in a molecule involved in desmosome formation and/or stability could be implicated in the pathophysiology of the disease.
|
Figure 1. Clinical features of DD. (a) Typical DD showing keratotic papules coalescing into a hyperkeratotic plaque on the chest. (b) Severe erosive DD with extensive fragility and fissuring on the thigh.
Linkage analysis in affected families mapped the DD locus to chromosome 12q23-24.1 with no evidence for locus heterogeneity (4-7). Recently, using a positional candidate approach, our group has established that mutations in the ATP2A2 gene cause DD (8). ATP2A2 (GenBank accession nos M23115 and M23114) encodes the sarco/endoplasmic reticulum Ca2+ ATPase isoform 2 (SERCA2). Alternative splicing of the ATP2A2 gene produces two isoforms, SERCA2a and SERCA2b, which differ in their C-termini and are expressed differently (9). SERCA2 belongs to a large family of P-type cation pumps including SERCA1 and SERCA3 (9,10). SERCA pumps play an important role in Ca2+ signal transduction by generating Ca2+ oscillations which mediate cell responses to extracellular signals. In non-excitable cells such as epithelium, the inositol 1,4,5-triphosphate (IP3)-mediated pathway predominates. After binding of ligands to plasma membrane receptors (G protein-coupled or tyrosine kinase receptors), IP3 acts as a second messenger and triggers release of Ca2+ from the endoplasmic reticulum (ER) (11). SERCA actively transports Ca2+ from the cytosol back into the ER lumen using energy from ATP hydrolysis (12). Repetitive cycles of discharge and uptake generate complex fluctuations in the intracellular concentration of Ca2+ which influence gene expression and cell differentiation. The efficiency and specificity of Ca2+ signalling is determined by the threshold, duration and frequency of the Ca2+ spikes which are regulated by the SERCA pumps (13-15).
The majority of the ATP2A2 mutations that we have reported in DD predict loss of function of the SERCA2 pump, suggesting that the dominant inheritance results from haploinsufficiency (8). Although it is not clear how loss of SERCA2 function causes DD, selective inhibition of SERCA pumps has been shown to interfere with the formation of intercellular junctions and cell-cell adhesion (16). In order to provide more information about the functional significance of ATP2A2 mutations, we have investigated another 21 families with DD as well as six sporadic cases of DD. We report 24 novel mutations in ATP2A2 and discuss the relationships between phenotype and genotype.
RESULTS
The entire coding sequence of ATP2A2 and flanking intron boundaries were amplified and screened for mutations by conformation-sensitive gel electrophoresis (CSGE) analysis in 21 families and six sporadic cases with DD. Bands showing abnormal electrophoretic mobility were identified in 16 families and in all of the sporadic cases. Sequencing of these bands led to the identification of 22 different mutations. Direct sequencing of the entire coding sequence of ATP2A2 in another four families revealed three missense mutations which could not be detected by CSGE analysis (Table 1). One of them is present in two families not known to be related. No mutation was found in two families with typical DD after direct sequencing of all the exons of ATP2A2.
Table 1. Novel ATP2A2 mutations in DD patients identified in this study
| Patients | Inheritance | Mutation | Nucleotide change | Consequence | Exon | Protein domain | Verification method |
| Nonsense (n = 4) | |||||||
| Ox-F8 | familial | Y122X | 366T->G | PTC | 5 | S2 | NlaIV |
| Ox-F27 | familial | R528X | 1582C->T | PTC | 13 | ATP binding | TaqI |
| Ox-F18 | familial | W551X | 1652G->A | PTC | 13 | ATP binding | CSCE |
| Ox-S25 | sporadic | W962X | 2898G->A | PTC | 20 | M10 | BspwI, MaeII |
| Frameshift insertion/deletion (n = 8) | |||||||
| Ox-S26 | sporadic | 130delG | PTC (+ 8 aa) | 2 | S1 | MboII | |
| Ox-F3 | familial | 1625delAG | PTC (+ 9 aa) | 13 | ATP binding | MboII | |
| Ox-F6 | familial | 2026insG | PTC (+ 2 aa) | 14 | ATP binding | BspWI | |
| Ox-F20 | familial | 2134delA | PTC (+ 43 aa) | 15 | hinge | SapI | |
| Ox-F5 | familial | 2608delAG | PTC (+ 3 aa) | 18 | M7-M8 loop | PleI, HinfI | |
| Ox-F14 | familial | 2675insC | PTC (+ 16 aa) | 18 | M8 | CSGE | |
| Ox-S38 | sporadic | 2703insCT | PTC (+ 2 aa) | 18 | M8 | CSGE | |
| Ox-S40 | sporadic | 2715delTA | PTC (+ 2 aa) | 18 | M8 | SfeI | |
| Splice site mutation (n = 1) | |||||||
| Ox-F29 | familial | 219+1G->A | exon skipping | 3 | S1-M1 | CSGE | |
| PTC (+ 4 aa) | |||||||
| In-frame insertion/deletion (n = 2) | |||||||
| Ox-F17 | familial | 137-12ins18 | in-frame ins. | 3 | S1 | NspI | |
| Ox-F33 | familial | 2258del3 | in-frame del. | 15 | S5 | CSGE | |
| Missense (n = 9) | |||||||
| Ox-S35 | sporadic | R131Q | 392G->A | missense | 5 | [beta]-strand | TaqI |
| Ox-S39 | sporadic | P160L | 479C->T | missense | 6 | [beta]-strand | CSGE |
| Ox-F10 | familial | S186P | 557C->T | missense | 7 | [beta]-strand | BsmAI |
| Ox-F21a | familial | C318R | 952T->C | missense | 8 | S4 | AciI |
| Ox-F9 | familial | E412G | 1235A->G | missense | 10 | phosphorylation | CSGE |
| Ox-F28 | familial | D702N | 2104G->A | missense | 15 | hinge | CSGE |
| Ox-F31 | familial | A745D | 2234C->A | missense | 15 | S5 | BbvI |
| Ox-F12 | familial | A803T | 2407G->A | missense | 16 | M6 | sequencing |
| Ox-F1&2 | familial | S920Y | 2759C->A | missense | 19 | M8-M9 loop | BsiYI |
Spectrum of mutations
Nonsense mutations.
We identified four base substitutions which resulted in premature termination codons (PTCs). Family Ox-F8 had a T->G transition at TAT tyrosine codon 122 and family Ox-F18 showed a G->A transition at TGG tryptophan codon 551. Both mutations lead to a TAG stop codon. Family Ox-F27 and patient Ox-S25 had a C->T transition at CGA arginine codon 528 and a G->A transition at TGG tryptophan codon 962, respectively, which resulted in a TGA stop codon.
Frameshift mutations.
Eight mutations create shifts in the reading frame. One sporadic case and one family showed a 1 bp deletion, 130delG and 2134delA, resulting in a stop codon eight and 43 codons downstream, respectively. Two families and one sporadic case had a 2 bp deletion, 1625delAG, 2608delAG and 2715delTA, leading to a stop codon nine, three and two codons downstream. Two families had a 1 bp insertion, 2026insG and 2675insC, which predicted a PTC two and 16 codons downstream, respectively. One patient exhibited a 2 bp insertion (2703insCT) which led to a PTC two codons downstream.
Splice site mutation.
One family (Ox-F29) showed a G->A transition at the invariant splice donor site consensus sequence GT of intron 3. The analysis of RT-PCR products from this patient revealed a smaller band in addition to the band of normal size. Direct sequencing of the smaller band revealed skipping of exon 3 which removes 83 nucleotides from G137 to T219 and leads to a frameshift resulting in a PTC three codons downstream of codon 46.
In-frame insertion and deletion.
Two families showed an in-frame insertion or deletion. Family Ox-F17 displayed an 18 bp insertion in intron 2 located 12 bp upstream of exon 3. Sequencing of the mutated transcript revealed abnormal splicing at a cryptic AG acceptor splice site located within the inserted sequence. This change predicts the addition of a new sequence (M48 F49 L50 T51 G52 K53) upstream of exon 3 in stalk region 1. Family Ox-F33 exhibited removal of an ACA triplet from an (ACA)3 repeat at nucleotide position 2258-2266 leading to the removal of asparagine codon 754 in stalk 5.
Mutations resulting in amino acid substitutions.
Nine single base changes leading to different amino acid substitutions were found in eight families and two sporadic cases. All of them occurred at highly conserved amino acid positions. None of these changes was detected in 50 normal controls using the verification methods described in Table 1. Mutation S920Y was present in two unrelated families (Ox-F1 and Ox-F2). Five mutations were located within cytoplasmic domains of the molecule. Mutations R131Q, P160L and S186P occurred within the [beta]-strand domain, mutation E412G is located in the phosphorylation domain and mutation D702N is in the hinge region which links the ATP-binding domain to the stalk region. Mutations C318R and A745D occurred within stalk domains 4 and 5, respectively. Mutation A803T was found within the transmembrane M6 domain which contains a calcium-binding site. Mutation S920Y occurred within the cytoplasmic loop between M8 and M9 (Fig. 2).
Figure 2. Localization and type of ATP2A2 mutations identified in patients with DD in this study. The predicted secondary structure of SERCA2 consists of 10 transmembranous domains and three globular cytoplasmic domains separated by a stalk sector. The cytoplasmic domains contain a [beta]-strand, a phosphorylation domain and an ATP-binding domain. The ATP-binding domain is linked to the stalk sector 5 by the hinge region. Four of the transmembrane domains (M4, M5, M6 and M8) contain Ca2+-binding sites (25). The 24 novel SERCA2 mutations identified in DD patients in this study are indicated. S1-S5 refer to stalk sectors and M1-M10 refer to transmembrane helices 1-10 (adapted from ref. 26 with permission).
During the search for mutations, a 2741+54 G->A polymorphism in intron 18 of ATP2A2 was detected in six families (Ox-F2, Ox-F6, Ox-F8, Ox-F12, Ox-F26 and Ox-F27). This change, which abolishes an MspI restriction enzyme site, was also found in six of 100 alleles in normal controls.
Haplotype analysis
To assess whether the S920Y mutation identified in families Ox-F1 and Ox-F2 resulted from independent events or from a common ancient mutation, haplotype analysis was performed by use of polymorphic microsatellite markers flanking the ATP2A2 gene (7). These two families shared a common disease-associated haplotype for all 16 markers from D12S2398 to D12S1332, suggesting that patients from these families had inherited the same mutation from a common ancestor. In contrast, the haplotypes associated with the disease for these microsatellite markers in the other families studied were all different. These results are consistent with the fact that the mutations identified in the other families are family specific and are likely to have occurred independently.
Genotype-phenotype comparison
In order to evaluate the functional significance of the ATP2A2 mutations, we analysed genotype-phenotype associations in 16 of the familial cases and one sporadic case investigated here and in five DD families described in our previous study (8). Skin lesions started around puberty or during young adulthood in most patients. All but one family (family Ox-F21) had classical DD in a `seborrhoeic pattern'. Keratotic papules and plaques were distributed over the scalp, chest, back, face and ears (Fig. 1a). All patients had some flexural disease. None of them had haemorrhagic lesions. At least one patient in each family showed palmar pits. Nail involvement, including longitudinal streaks and notches, was present in most individuals. In some patients, the disease was exacerbated by various factors including stress, UV exposure, heat and sweat, friction and oral contraceptive pills. Bacterial infections were noted in some patients. The response to oral retinoids and topical steroids was variable. Painful erosive lesions and extensive flexural involvement suggestive of Hailey-Hailey disease dominated the clinical picture in two members of family Ox-F21 (Fig. 1b). However, these patients also had keratotic papules, palmar pits and nail changes characteristic of DD. Three patients had epilepsy (families Ox-F1, Ox-F6 and Ox-F15) and three patients had personality problems which resulted in antisocial behaviour (families Ox-F12, Ox-F15 and Ox-F18). One patient (family Ox-F6) had bipolar disorder. One patient (family Ox-F15) developed spinocerebellar ataxia.
Mutations were categorized into three groups according to their predicted consequence on the SERCA2 protein (Table 2). Marked inter- and intrafamilial phenotypic variations were noted in each group. Progression of the disease also seemed extremely variable. The first group consists of 10 families and one sporadic case, each with distinct mutations leading to a PTC. These mutations predict the absence of protein synthesis through nonsense mRNA decay (17). None of this group (22 patients) had severe DD. The second group is made up of four families with in-frame mutations predicting the synthesis of a shortened or elongated SERCA2 protein. One patient from family Ox-F15 had severe disease but other affected members of this family had a milder phenotype. The other three families had mild (five patients) or moderate (five patients) disease. The third group comprises seven families in which six distinct mutations resulted in non-conservative amino acid substitutions. Patients from families Ox-F1 and Ox-F2 share the same missense mutation (S920Y), one patient from each family having severe disease with widespread skin lesions including hypertrophic flexural lesions and oral mucosal involvement. Other members of these two families have milder symptoms. Two individuals from family Ox-F21 who carry the C318R mutation show severe erosive or bullous skin lesions (Fig. 1b). The rest of the families in this group present mild (two patients) or moderate (six patients) disease. In each group of mutations, some patients from a total of five different families had neuropsychiatric manifestations (Table 2).
Table 2. Genotype-phenotype correlation in patients whose disease phenotype could be assessed precisely
| Patients | Protein domain | Age of onseta (n) | Severity (n) | Oral lesions (n) | Progression (n) | Neuropsychiatric manifestations (n) |
| Group 1. Nonsense and frameshift mutation (n = 11) | ||||||
| Ox-F29 (2) | S1-M1 | b (2) | mild (1), moderate (1) | yes (1) | improve (2) | no |
| Ox-F8 (2) | S2 | b (2) | moderate (2) | no | improve (1), static (1) | no |
| Ox-F18 (2) | ATP binding | b (2) | mild (2) | no | static (1), worse (1) | behaviour problem (1) |
| Ox-F3 (2) | ATP binding | a (1), c (1) | moderate (2) | no | improve (2) | no |
| Ox-F6 (3) | ATP binding | a (1), b (2) | moderate (3) | no | improve (1), static (2) | epilepsy (1) |
| bipolar disorder (1) | ||||||
| Ox-F20 (1) | hinge | b | moderate (1) | no | no data | no |
| Ox-F5 (5) | M7-M8 loop | a (2), b (2), c (1) | mild (4), moderate (1) | yes (1) | static (1), worse (1), no data (3) | no |
| Ox-F14 (1) | M8 | b | moderate (1) | no | improve (1) | no |
| Ox-S40 | M8 | b | moderate | no | improve | no |
| Ox-F13b (2) | M9 | b (2) | mild (1), moderate (1) | no | static (2) | no |
| Ox-F27 (1) | M10 | d | moderate (1) | no | static (1) | no |
| Group 2. In-frame insertion/deletion and exon skipping (n = 4) | ||||||
| Ox-F15 b (5) | before S1 | b (5) | mild (1), moderate (3), severe (1) | no | improve (1), static (2), worse (1) | epilepsy (1) |
| behaviour problem (1) | ||||||
| ataxia (1) | ||||||
| Ox-F22b (1) | hinge | c | mild (1) | yes | worse (1) | no |
| Ox-F33 (3) | S5 | a (1), b (2) | mild (3) | no | improve (1), worse (1) | no |
| Ox-F4b (2) | phosphorylation | b (1), c (1) | moderate (2) | yes (1) | static (1), worse (1) | no |
| Group 3. Missense (n = 6) | ||||||
| Ox-F21c (2) | S4 | a (2) | severe (2) | yes (2) | static (1), worse (1) | no |
| Ox-F11b (1) | phosphorylation | d | moderate | no | no data | no |
| Ox-F28 (1) | hinge | b | moderate | no | worse (1) | no |
| Ox-F31 (3) | S5 | a (1), b (1) | moderate (3) | no | static (1), worse (2) | no |
| Ox-F12 (3) | M6 | a (2), b (1) | mild (2), moderate (1) | no | static (3) | behaviour problem (1) |
| Ox-F1 (5) | M8-M9 loop | b (5) | mild (1), moderate (3), severe (1) | yes (2) | static (2), worse (2), no data (1) | epilepsy (1) |
| Ox-F2 (3) | M8-M9 loop | b (3) | moderate (2), severe (1) | yes (3) | improve (1), static (1), worse (1) | no |
aAge of onset, a < 10 years; b = 10-20 years; c = 20-40 years; d > 40 years of age.
bPatients whose mutation was reported previously (8).
cErosive form of DD.
DISCUSSION
We have identified a total of 24 novel mutations in the ATP2A2 gene in 19 families and six sporadic cases with DD. Thirteen mutations (54%) predict premature termination of translation, two mutations (8%) result in in-frame insertion or deletion and nine mutations (38%) predict amino acid substitutions. All mutations except one (S920Y) are family specific.
Predicted effects of ATP2A2 mutations on SERCA2 function
Fifty-four per cent of the ATP2A2 mutations identified in this study are nonsense mutations or small deletions or insertions causing frameshifts leading to PTCs. These mutations predict nonsense-mediated mRNA decay (17) and absence of production of the mutated SERCA2 polypeptide. These findings support the proposition that haploinsufficiency (i.e. production of a normal phenotype requires more gene product than produced by a single copy of the gene) is a common mechanism for the dominant inheritance of DD.
A smaller percentage (38%) of ATP2A2 mutations identified in this study are missense mutations leading to non-conservative amino acid substitutions. These mutations occur at positions which are highly conserved among species and other genes in the same family. These changes have not been found in 50 normal individuals, and all segregated with the disease in the DD families. Thus it is unlikely that they represent silent polymorphisms. Four mutations are located at positions which have been studied by site-directed mutagenesis of SERCA1 (11,18-20). Mutations D702N, A803T and P160L occur at nucleotide positions that have been shown to result in total or partial loss of Ca2+ transport activity (11,18,19). Mutation C318R changes a hydrophobic amino acid into a basic one in stalk 4 of the molecule. Mutation of the same residue to an alanine was shown to have no effect on Ca2+ transport (20). However, the function of SERCA2 in keratinocytes has not been studied previously and the functional importance of specific SERCA2 domains in these cells may be different. As all these missense mutations introduce non-conservative changes which alter the charge, polarity, hydrophobicity and/or size of the corresponding amino acid residue in functional domains of the molecule, it is likely that they will interfere with SERCA2 function. Functional studies of these mutations are required, together with studies of the effect of the small in-frame deletion (family Ox-F17) and insertion (family Ox-F33) occurring in stalk segments 1 and 5.
Our results suggest that epidermal cells are sensitive to ATP2A2 gene dosage and that SERCA2 is a critical contributor to normal homeostasis in the adult epidermis and in some areas of the body. In general, DD presents around puberty; the cutaneous lesions have a focal distribution which predominates in seborrhoeic areas and the disease is exacerbated by heat, sweating and the sun. We suggest that homeostatic mechanisms in the epidermis can compensate for subtle abnormalities in Ca2+ signalling but local triggers reveal the abnormality in DD. Compensatory mechanisms may include increased expression of the normal ATP2A2 allele and/or compensation by other SERCA pumps expressed in the epidermis such as SERCA1 and SERCA3. Alternatively, the activity of SERCA2 pumps required in different cutaneous areas may vary depending on physiological and/or external factors.
Although the majority of ATP2A2 mutations identified in this study predict the loss of function of the mutated SERCA2 polypeptide, we cannot exclude the possibility that some ATP2A2 mutations could act through mechanisms distinct from haploinsufficiency. In particular, some missense or in-frame ATP2A2 mutations may have residual or abnormal function. Overexpression of ER calcium pumps in Xenopus oocytes alters the frequency of Ca2+ oscillations (21), suggesting that SERCA2 function could be sensitive to increased gene dosage. It is also possible that SERCA2 may interact with potential regulatory proteins and that some mutations impair regulation of SERCA2 function by these molecules. Functional analyses of ATP2A2 mutations which do not predict absence of protein synthesis are required to test these possibilities.
In two patients with classical DD, no mutation was found in the ATP2A2 gene despite sequencing the entire coding region and intronic splice sites of the gene. Southern blot analysis of these patients' genomic DNA using MboI and TaqI restriction enzymes and an ATP2A2 probe failed to reveal a gross rearrangement (data not shown). In the absence of evidence for genetic heterogeneity from linkage studies (4-7), it is possible that these patients harbour mutations in an unscreened region of the ATP2A2 gene such as the promoter region, intronic sequences or 3[prime]-untranslated region which could affect the expression and/or function of the SERCA2 protein.
Distribution of ATP2A2 mutations
The mutations were spread throughout the ATP2A2 gene. No clustering or hot spots of mutations have been observed. In addition, no mutation was located in the C-terminal region of the SERCA2 protein which is specific for the alternatively spliced isoforms SERCA2a and SERCA2b (see Table 3 footnotes). In contrast to mutations leading to PTCs which are expected to result in null alleles, missense and in-frame mutations are predicted to affect specific domains of the SERCA2 pump. Ca2+ transport from the cytosol to the ER lumen involves critical interactions between cytoplasmic and transmembrane domains of the SERCA2 pump. Upon binding of two Ca2+ ions to transmembrane domains M4, M5, M6 and M8, phosphorylation from ATP results in conformational changes of the SERCA2 protein from a high to a low Ca2+ affinity phosphoenzyme intermediate (11). These changes involve long-range interactions through stalk 4 and 5 sectors and result in the release of the two Ca2+ ions into the ER lumen. Of the missense and in-frame mutations that we report, six are located within the cytoplasmic [beta]-strand, the phosphorylation domain, the ATP-binding domain or the hinge region, and four mutations occurred within the stalk and transmembrane domains of the SERCA2 molecule. It is thus possible that these non-conservative changes impair specific interactions between functional domains of the molecule. Functional studies of the mutations which have not been studied previously by site-directed mutagenesis are needed to help in understanding their specific effect(s) on SERCA2 function. Mutation S920Y occurred within the cytoplasmic loop between M8 and M9, a region whose function is unknown. The observation that members of both families who shared this mutation have severe clinical manifestations suggests that this region plays an important role in SERCA2 function.
Table 3. PCR primers for amplification of ATP2A2 from genomic DNA
| Exon | Primer (5[prime]->3[prime]) | Position of primer | Product size | Annealing |
| in the intron | (bp) | temperature (C°) | ||
| 1 | CGAGGCGGAGGCGAGGAG | -60a | 271 | 55 |
| GGAGCCGAAGCCCACGCG | +93 | |||
| 2 + 3 | ACCTCCCTCTTGACACATTG | -45 | 464 | 55 |
| GACAACTCCTAACCACACTG | +227 | |||
| 4 | CGTGCCATTTCTCTTCTAGG | -20 | 222 | 55 |
| CTCAACACATCAGGAAAAACAG | +98 | |||
| 5 | AGTGTCAGGCAGGTCTTTAC | -45 | 368 | 55 |
| AGGAAGGGAGGTGCTAAAAC | +184 | |||
| 6 | AGCCTCATTCTCTTCCTTCC | -224 | 455 | 55 |
| ATGGAGCGAGACTAAAGCAC | +150 | |||
| 7 | CTTGGTGTGGGTCGCAGAG | -50 | 238 | 50 |
| CCTTTAGAATGATAGCCAGTG | +102 | |||
| 8 | GTTGTATGGCTGGTTGCTTG | -147 | 657 | 55 |
| GAACAAAGAACCACGACACG | +45 | |||
| 9 | GGTTGTTTGCCTTTGTCCTAA | -98 | 426 | 50 |
| ATAACAAACACAAATCCCTCTT | +239 | |||
| 10 | GGCGACCATACCCTGCTC | -42 | 202 | 55 |
| CCCACCCCACCCTTGAAC | +57 | |||
| 11 | TCAGAGGAGGATAAAAATGGC | -136 | 389 | 50 |
| CTGTAAGTTTGAGGAGATAAGG | +121 | |||
| 12 + 13 | ATTGCCACCCAGTAGTATCC | -39 | 568 | 55 |
| GAACTGTTTGACCTTTTGCTTG | +88 | |||
| 14 | CTAGAACTTGCCACTTTTATTTA | -43 | 436 | 50 |
| GAGGCTACTATGTGCTTGTG | +57 | |||
| 15 | TTTCCTCCTGCTTCCCATTC | -91 | 400 | 55 |
| GCAATCTGGAGAGCAAACTG | +88 | |||
| 16 | TCATTTATTTTTCTGGAGGAGG | -100 | 430 | 55 |
| CATCTCTGTCTTTTGCTACCC | +127 | |||
| 17 | TGATCTTCGTCCTTGTGGGG | -94 | 261 | 55 |
| TGATAGATACCGAAACCACAG | +81 | |||
| 18 | GGGTTGGAGCCTGGACTTG | -65 | 312 | 55 |
| TTTTGGGAAGGGAAGAACTGT | +114 | |||
| 19 + 20 | TCCCCACCTCTCCTTGCTC | -24 | 609 | 55 |
| CCTCCATCACCAGCCAGTAT | +115b | |||
| 21c | GTTCCTTTTCATCTGTCGCTG | -105 | 228 | 50 |
| TCTTTTTCCCCAACATCAGTC | +109b |
aNucleotide position 5[prime] to the ATG start codon.
bNucleotide position 3[prime] to the stop codon.
cPrimer set for exon 21 is specific for isoforms ATP2A2a. [The sequence located in exon 20 is spliced at a GT cryptic donor splice site to generate the ATP2A2a isoform by joining exon 21 to the first 121 nucleotides of exon 20. The ATP2A2b isoform contains the entire exon 20 and no exon 21 sequence (10).]
Genotype-phenotype correlations
When we categorized the mutations according to the predicted effect on the mutated protein (absence of protein synthesis versus production of a structurally abnormal polypeptide), it appeared that severe disease phenotype was seen mainly in patients carrying a missense or an in-frame insertion (Ox-F1, Ox-F2, Ox-F15 and Ox-F21), including one with erosive DD (Ox-F21) (Table 2). In addition, a higher percentage (44%; 11/25) of patients carrying a missense or an in-frame insertion or deletion experienced progression of disease severity with age in comparison with patients whose mutation led to a PTC (11%; 2/18). Longer follow up is required to confirm these observations. However, it is interesting to note that in a recent study, missense mutations in ATP2A2 were often associated with atypical or severe clinical features (22). Finally, there were considerable phenotypic variations within and between the families that we studied, suggesting that modifying genes and/or environmental factors influence the phenotype. The neuropsychiatric abnormalities identified in our study were not associated with a specific type of mutation and were not constant among affected members within the same family. Therefore, the interpretation of these findings requires genetic and phenotypic investigation of additional families.
Although ATP2A2a is highly expressed in cardiac, smooth and slow twitch skeletal muscles, diseases affecting these tissues do not appear to be increased in our patients with DD. Interestingly, heterozygote knock-out mice for ATP2A2 display reduced cardiac performance despite the absence of obvious clinical phenotype (23). The relevance of these findings to DD is unknown. Further studies are needed to investigate the molecular mechanisms of the disease as well as modifying factors which influence the disease phenotype. These results will provide the basis for new therapies.
MATERIALS AND METHODS
Patients
We studied 21 families (75 individuals) and six sporadic cases with DD. All families and four of the sporadic cases were British. One sporadic case was French and one was Italian. The diagnosis was made by dermatologists based on clinical and histopathological findings.
History and clinical findings were recorded in 21 families and in one sporadic case who had been examined by one dermatologist (S.B.) (1). The general medical history was recorded including neuropsychiatric and cardiac disease. The following clinical parameters were assessed: age of onset, pattern of disease (including hand and nail involvement), exacerbating factors, responsiveness to treatment, susceptibility to infection and progression of the disease with age, and associated findings were noted. The disease in adults (age >25 years) was classified as mild, moderate or severe. Patients having mild disease developed keratotic papules scattered sparsely over the trunk or flexures or disease limited to one or two areas. Patients who had more extensive papular lesions or localized verrucous plaques were classified as moderate cases. Patients with coalescent verrucous plaques involving most of the trunk or grossly hypertrophic flexured disease were considered as having severe disease. ATP2A2 mutations in five of these families have been reported previously (Table 2) (8).
PCR and RT-PCR amplification
Genomic DNA was extracted from peripheral blood leukocytes using standard methods. PCR amplification of ATP2A2 was performed from genomic DNA with 19 sets of primers spanning all 21 exons and flanking splice sites of the gene (Table 3). In 20 patients, total RNA was extracted from transformed lymphoblasts using Trizol (Life Technologies). Reverse transcription was performed using random hexanucleotides (Pharmacia Biotech) and Superscript reverse transcriptase (Life Technologies) at 37°C for 45 min. The entire coding sequence of ATP2A2 cDNA was amplified by means of RT-PCR reactions using eight sets of primers (Table 4). The conditions for PCR amplification consisted of 94°C for 4 min followed by 35 cycles of 94°C for 30 s, either 55°C or 50°C for 30 s, and 72°C for 30 s in an MJ Research PTC-100 thermal cycler. All PCRs were performed in buffer containing 1.5 mM MgCl2, 45 mM Tris pH 8.8, 11 mM (NH4)2SO4 pH 8.8, 6.7 mM [beta]-mercaptoethanol and 4.5 µM EDTA. All PCR products were run on 2% agarose gels to check for the presence of a large insertion, deletion or splicing abnormalilty before analysis by CSGE.
Table 4. Primers for RT-PCR amplification of ATP2A2 cDNA
| Primer set | Exons amplified | Forward (5[prime]->3[prime]) | Reverse (5[prime]->3[prime]) | Product size (bp) | Annealing temperature (°C) |
| 1 | 1-6 | CGAGGCGGAGGCGAGGAG | GTGTGGTAGATTTGATGGAAG | 577 | 55 |
| 2 | 5-8 | GAGTGTGCAGCGGATTAAAG | GCAATGCAAATAAGGGAGATG | 405 | 55 |
| 3 | 8-10 | CAACAGAACAGGAGAGAACAC | CAGAGTCATTACAAAGAGCAC | 549 | 55 |
| 4 | 9-13 | AACTGGATCAACTTATGCACC | CTGCCACTACCCCACTCTC | 513 | 55 |
| 5 | 13-14 | ATGTTTGTGAAGGGTGCTCC | GTTCACGCCATCGCCAGTC | 585 | 55 |
| 6 | 14-16 | ACGCCCGCTGTTTTGCTCG | CAGCCAATAGCCAAGTAACG | 511 | 55 |
| 7 | 16-20 | GCTGCTCTGGGTCAATCTG | GGCAAGGAGATTTTCAGCAC | 555 | 55 |
| 8hk1a | 19-20 | CTTCCTGATCCTCTATGTCG | ACAATGTCTGCTGGCTCAAC | 625 | 55 |
| 8hk2a | 19-21 | CTTCCTGATCCTCTATGTCG | ACGCAACCGAACACCCTTAC | 213 | 55 |
CSGE analysis
After PCR amplification, the products were heated at 98°C for 5 min, followed by 68°C for 1 h to induce heteroduplex formation. The PCR products were loaded on a gel containing 10% acrylamide, 15% formamide, 10% ethylene glycol and 0.5% glycerol-tolerant buffer (USB) and were subjected to electrophoresis at 500 V for 16 h (24).
Sequence analysis
RT-PCR and PCR products showing abnormal electrophoretic mobility were purified using Qiaquick PCR purification columns (Qiagen). PCR products which showed multiple bands on 2% agarose gel electrophoresis were gel purified on 2% low melting agarose gels using Qiaex gel purification kit (Qiagen). The purified products were sequenced directly using dideoxy terminator cycle sequencing and an Applied Biosystems model 373A automated sequencer. All products were sequenced in the forward and reverse orientation. Some of the products were subcloned into the pGEMT vector (Promega) and the plasmid insert was sequenced. The entire coding region of the ATP2A2 gene and intronic splice sites were sequenced in patients whose PCR-amplified products showed a normal migration pattern by CSGE analysis.
Haplotype analysis
Haplotyping was performed by PCR amplification of genomic DNA from patients and other members of the families using polymorphic microsatellite markers flanking the ATP2A2 gene from marker D12S1339 to D12S129). Products, with an equal amount of formamide buffer, were separated on denaturing 6% polyacrylamide gels (National Diagnostics) and were transferred onto a Biodyne B nylon membrane (Pall). The membranes were hybridized with a primer that was [[alpha]-32P]dCTP end-labelled (3000 Ci/mmol; Amersham) by means of terminal transferase (Boehringer Mannheim) and were autoradiographed on X-ray film (X-Omat-AR; Kodak).
ACKNOWLEDGEMENTS
We are grateful to the clinicians and the families who participated in the study. A.S. has a fellowship from the Royal Thai Government and the Faculty of Medicine, Ramathibodi Hospital, Mahidol University, Bangkok; S.M. held a Wellcome Trust prize studentship. A.H. is a recipient of a DEBRA UK research fellowship.
REFERENCES
+To whom correspondence should be addressed. Tel: +44 1865 287511; Fax: +44 1865 287501; Email: alain.hovnanian{at}well.ox.ac.uk
This page is run by Oxford University Press, Great Clarendon Street, Oxford OX2 6DP, as part of the OUP Journals
Comments and feedback: jnl.info{at}oup.co.uk
Last modification:
Copyright© Oxford University Press, 1999.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  Y. Miyauchi, T. Daiho, K. Yamasaki, H. Takahashi, A. Ishida-Yamamoto, S. Danko, H. Suzuki, and H. Iizuka Comprehensive Analysis of Expression and Function of 51 Sarco(endo)plasmic Reticulum Ca2+-ATPase Mutants Associated with Darier Disease J. Biol. Chem., August 11, 2006; 281(32): 22882 - 22895. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B M Mayosi, A Kardos, C H Davies, F Gumedze, A Hovnanian, S Burge, and H Watkins Heterozygous disruption of SERCA2a is not associated with impairment of cardiac performance in humans: implications for SERCA2a as a therapeutic target in heart failure Heart, January 1, 2006; 92(1): 105 - 109. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. Sato, K. Yamasaki, T. Daiho, Y. Miyauchi, H. Takahashi, A. Ishida-Yamamoto, S. Nakamura, H. Iizuka, and H. Suzuki Distinct Types of Abnormality in Kinetic Properties of Three Darier Disease-causing Sarco(endo)plasmic Reticulum Ca2+-ATPase Mutants That Exhibit Normal Expression and High Ca2+ Transport Activity J. Biol. Chem., August 20, 2004; 279(34): 35595 - 35603. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. Dode, J. P. Andersen, N. Leslie, J. Dhitavat, B. Vilsen, and A. Hovnanian Dissection of the Functional Differences between Sarco(endo)plasmic Reticulum Ca2+-ATPase (SERCA) 1 and 2 Isoforms and Characterization of Darier Disease (SERCA2) Mutants by Steady-state and Transient Kinetic Analyses J. Biol. Chem., November 28, 2003; 278(48): 47877 - 47889. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. Hulatt and S. Burge Darier's disease: hopes and challenges J R Soc Med, September 1, 2003; 96(9): 439 - 441. [Full Text] [PDF]
|
|
|
|
|
|
W. Ahn, M. G. Lee, K. H. Kim, and S. Muallem Multiple Effects of SERCA2b Mutations Associated with Darier's Disease J. Biol. Chem., May 30, 2003; 278(23): 20795 - 20801. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. Sudbrak, J. Brown, C. Dobson-Stone, S. Carter, J. Ramser, J. White, E. Healy, M. Dissanayake, M. Larregue, M. Perrussel, et al. Hailey-Hailey disease is caused by mutations in ATP2C1 encoding a novel Ca2+ pump Hum. Mol. Genet., April 12, 2000; 9(7): 1131 - 1140. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
V. L. Ruiz-Perez, S. A. Carter, E. Healy, C. Todd, J. L. Rees, P. M. Steijlen, A. J. Carmichael, H. M. Lewis, D. Hohl, P. Itin, et al. ATP2A2 mutations in Darier's disease: variant cutaneous phenotypes are associated with missense mutations, but neuropsychiatric features are independent of mutation class Hum. Mol. Genet., September 1, 1999; 8(9): 1621 - 1630. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||