Acral peeling skin syndrome with TGM5 gene mutations may resemble epidermolysis bullosa simplex in young individuals.

TO THE EDITOR The acral peeling skin syndrome (APSS) is a rare autosomal recessive condition characterized by superficial painless peeling of the skin predominantly on the dorsal aspects of hands and feet (Shwayder et al., 1997; Cassidy et al., 2005). The condition is usually aggravated by heat, humidity, and exposure to water. Microscopically, the cleavage level is located in the upper epidermis, between the stratum granulosum and the stratum corneum (Garcia et al., 2005). Only 15 patients with APSS have been reported since 1997 (Shwayder et al., 1997; Brusasco et al., 1998; Hashimoto et al., 2000; Cassidy et al., 2005; Garcia et al., 2005; Kharfi et al., 2009; Oumakhir et al., 2009; Wakade et al., 2009) (Table 1). In several of them, in addition to superficial peeling, acral blisters were also described (Wakade et al., 2009). The genetic basis of the disease was determined in only three families, in whom two different missense mutations in the TGM5 gene encoding transglutaminase 5 (TGase 5) were disclosed (Table 1) (Cassidy et al., 2005; Kharfi et al., 2009). It remains unclear whether the other patients have mutations in the same gene, or whether APSS is clinically and genetically heterogeneous (Cassidy et al., 2005). In this study, we investigated nine unrelated patients, eight children and one adult, clinically suspected to have epidermolysis bullosa simplex (EBS) because of acral skin blistering. The patients and/or diagnostic samples were referred to the Epidermolysis bullosa Center of the University Medical Center Freiburg (Volz et al., 2007) for molecular diagnostics of EBS. EDTA-blood and skin samples were obtained after informed consent of the patients and, if available, of family members. EDTA-blood samples of 50 clinically unaffected Central European individuals were used as controls. The study was conducted according to the Declaration of Helsinki Principles. Immunofluorescence staining of skin cryosections was performed using a panel of antibodies to components of the epidermal basement membrane zone (Kern et al., 2006), as well as antibodies to loricrin (Abcam, Cambridge, UK), filaggrin (clone 15C10; Novocastra, Newcastle, UK), involucrin (clone SY5; Sigma, Taufkirchen, Germany), cytokeratin 10 (clone DE-K10; Dako, Glostrup, Denmark), TGase 1 (clone B.C1; Biomedical Technologies, Madrid Spain), TGase 3 (Jackson Immunoresearch Laboratories, West Grove, PA), and TGase 5 (Novus Biologicals, Littleton, CO). Genomic DNA was extracted from EDTA-blood using the QiAmp DNA mini kit (Qiagen, Hilden, Germany). Amplification of all KRT5 (NC_000012.11, National Center for Biotechnology Information (NCBI)), KRT14 (NC_000017.10, NCBI), and TGM5 (NC_000015.9, NCBI) exons and exon–intron boundaries, and sequencing were performed as described (Schuilenga-Hut et al., 2003; Wood et al., 2003; Cassidy et al., 2005). Mutations were confirmed by resequencing. The mutation c.763T>C was verified in 100 control chromosomes by digestion with the restriction enzyme BsrBI according to the manufacturer's protocol (NEBioLabs, Ipswich, MA). Predictions regarding the consequences of the mutation and modeling were performed using the software Polyphen (http://coot.embl.de/PolyPhen/) and SWISS-MODEL (http://swissmodel.expasy.org/workspace/index.php?func=modelling) (Guex and Peitsch, 1997). The clinical phenotypes of the nine patients in this study are summarized in Table 1. It is noteworthy that eight of the nine patients are children of 12 years of age or younger, with only one adult in the cohort. Eight had had blisters at birth or since infancy (Figure 1a); for patient 6, no information was available. Healing occurred spontaneously, sometimes with residual erythema, burning sensation, or pruritus, but without scarring or atrophy. Aggravation was caused by sweating, heat, and humidity, or by mechanical trauma. In most cases, only the volar and dorsal aspects of hands and feet were affected, but in the 47-year-old patient 4, exfoliation also occurred on elbows and knees (Figure 1c). In older children and in the adult patient, peeling of the skin was the most prominent symptom, but blisters occurred occasionally. Interestingly, three of the patients reported having a parent (patients 5 and 9) or offspring (patient 4) with mild features of the disease. In the skin samples of the patients, there were no abnormalities of the basement membrane zone, and all markers stained positive. This result is compatible with the diagnosis of localized EBS, because in these patients fragility of the skin is not prominent. In our experience, in the biopsy samples sent for immunofluorescence mapping to the Epidermolysis bullosa Center Freiburg, tissue separation is observed only in about a third of the localized EBS cases. Furthermore, keratin staining is usually not altered. In some samples in the present cohort, a discrete split in the stratum corneum was present (Figure 2), suggesting the diagnosis of APSS, which was thereafter confirmed by mutation analysis. Mutations in the exons and intron/exon boundaries of the KRT5 and KRT14 genes were excluded in all patients. In patients 2–9, the TGM5 mutation c.337G>T, p.Gly113Cys was disclosed in a homozygous state. As reported before (Cassidy et al., 2005), the mutation was associated with the single-nucleotide polymorphism p.Thr109Met, also in a homozygous state. The father of patient 5 and the parents of patient 7 were heterozygous carriers of the mutation and single-nucleotide polymorphism, respectively. Patient 1 was compound heterozygous for p.Gly113Cys and for the mutation c.763T>C, p.Trp255Arg (Figure 2e). To our knowledge, this mutation is previously unreported and was excluded among 100 control chromosomes by restriction digestion with BsrBI. Tryptophan 255 is conserved in all TGases in different species and lies within the core domain close to the catalytic active site of TGase 5 (see Supplementary Figure S1 online). The three-dimensional structure of the core domain of the wild-type and Trp255Arg-mutated TGase 5 was modeled using the closely related TGase 3 as a template (PDB ID, 1I9mA; chain ID, 4) (Grenard et al., 2001). This model revealed proximity of amino-acid 255 to the catalytic site of the enzyme and the conformation changes induced by the amino-acid substitution (Supplementary Figure S1). Because the mutation substitutes the large nonpolar neutral tryptophan with the smaller, polar and positively charged arginine, we performed a prediction of electrostatic potentials using the Poisson–Boltzmann equation. This revealed significant differences in electrostatic potentials between wild-type (Trp255) and mutated (Arg255) core domains (Supplementary Figure S1). Significantly, the amino-acid stretch Ser492–Ser501, which has been recently described as an important cleavage region for proteolytic activation of TGase 5 (Pietroni et al., 2008), was strongly affected by conformation and electrostatic changes. Therefore, we predict that the mutation has a considerable impact on TGase 5, inducing conformation and electrostatic changes. Compared to wild type, the mutated TGase 5 was restricted to a few cell layers. Similar to TGase 1 (Raghunath et al., 1998), missense mutations may alter activity but not enzyme expression. In APSS skin, TGase 1 staining was diffuse, but TGase 3 was not changed (Figure 2). Kinetic and in vitro experiments have indicated that TGase 5 is very efficient in cross-linking loricrin, involucrin, and small proline-rich proteins (Candi et al., 2001). Therefore, we investigated the effect of TGase 5 mutations on the distribution of loricrin, involucrin, and filaggrin in the skin of five patients in vivo (1–5). Interestingly, all three proteins were distributed in a more diffuse manner in patients' skin samples compared to controls (Figure 2). However, the molecular mechanisms behind these findings remain to be elucidated. Taken together, we report that patients with clinically suspected EBS carried TGM5 mutations and in fact suffered from APSS. Such an initial assumption of epidermolysis bullosa has been reported in literature before (Hashimoto et al., 2000) and is understandable for several reasons: first, in infants, APSS frequently manifests with blistering on palms and soles and is aggravated by mechanical factors and by a humid, warm environment. Second, in three families, a vertical transmission of the disease was considered because either parents or offspring were reported to have had skin blistering in childhood. Third, similar to that observed occasionally in localized EBS, no tissue separation was detected by light microscopic examination of skin biopsies. The superficial skin split visible in some skin sections was considered to be an artifact in cryosections. For these reasons, the diagnosis of APSS was missed. Moreover, the presumption that APSS affects only the dorsal regions of the hands and feet (Cassidy et al., 2005) might have confused some clinicians. On the basis of the above, we consider APSS to be an important differential diagnosis for EBS, which certainly accounts for a number of cases in which keratin mutations were not identified. Indeed, of 83 individuals referred to us with suspected EBS, 11% had TGM5 mutations and APSS. Importantly, the TGM5 mutation p.Gly113Cys was identified in nine unrelated patients, which suggests that it is a recurrent, possibly an ancestral, mutation in the European population (Cassidy et al., 2005). This finding has practical consequences for diagnostic testing in patients suspected to have EBS but no keratin mutations. In such cases, screening of the recurrent TGM5 mutation p.Gly113Cys is easy and rapid. Finally, molecular genetics will shed light on the different APSS phenotypes. On the basis of the knowledge on related skin diseases such as epidermolysis bullosa, genetic and phenotypic heterogeneity should be expected. The authors state no conflict of interest. We thank all patients who participated in this study and the physicians who sent us samples. The excellent technical support by Gabriele Grüninger, Vera Morand, and Kaethe Thoma is gratefully acknowledged. This work was supported by the Network Epidermolysis Bullosa grant from the Federal Ministry for Education and Research (BMBF) to LBT, the Excellence Initiative of the German Federal and State Governments and the Freiburg Institute for Advanced Studies, FRIAS, School of Life Sciences to LBT, and by the K Kriezis scholarship from the National and Kapodistrian University of Athens to DK. The study protocol was approved by the ethics committee of the University Medical Center Freiburg. SUPPLEMENTARY MATERIAL Supplementary material is linked to the online version of the paper