Content » Vol 94, Issue 3

Clinical Report

T-lymphocytes are Directly Involved in the Clinical Expression of Migratory Circinate Erythema in Epidermolysis Bullosa Simplex Patients

Daniele Castiglia1, May El Hachem2, Andrea Diociaiuti2, Teresa Carbone3, Naomi De Luca1, Monica Pascucci1, Giovanna Zambruno1 and Andrea Cavani3

1Laboratory of Molecular and Cell Biology, 3Laboratory of Experimental Immunology, Istituto Dermopatico dell’Immacolata-IRCCS, and 2Dermatology Division, Ospedale Pediatrico Bambino Gesù-IRCCS, Rome, Italy

Epidermolysis bullosa simplex with migratory circinate erythema (EBS-MCE) is a rare EBS subtype characterised by migratory blistering lesions that resolve with brownish pigmentation. It is caused by a recurrent readthrough mutation, c.1649delG, in the tail of keratin 5. Here, we report a child with EBS-MCE and investigated the immunologic mechanisms underlying the migratory lesions in this patient. A skin biopsy from the patient from an active border of an erythematous lesion was used for the immunohistochemical characterisation of the inflammatory infiltrate and for TUNEL assay to detect apoptotic cells. We found abundant CD4+ and CD8+ T lymphocytes infiltrating the papillary dermis and lining the dermal–epidermal junction. A number of these cells expressed the activation marker CD69. CD83+ dendritic cells were present both in the epidermis and papillary dermis. Finally, TUNEL staining showed apop­tosis of basal and suprabasal keratinocytes. These findings suggest a critical role of the cellular immunity in determining the EBS-MCE phenotype. Key words: genodermatosis; intermediate filaments; mottled pigmentation; nonhelical tail domain.

Accepted May 21, 2013; Epub ahead of print Oct 8, 2013

Acta Derm Venereol 2013;

Daniele Castiglia, PhD, Laboratory of Molecular and Cell Biology, Istituto Dermopatico dell’Immacolata-IRCCS, via dei Monti di Creta 104, IT-00167 Rome, Italy. E-mail:

Epidermolysis bullosa simplex (EBS) is a group of inherited skin diseases characterised by intraepidermal blistering. According to the level of blister formation, EBS is distinguished in the basal and suprabasal type. The latter refers to less common phenotypes associated with split in the suprabasal epidermal layers (1). Basal EBS (hereafter defined as EBS) is the classical type, which results from cytolysis of basal keratinocytes and cell fragility throughout the basal layer upon mechanical trauma. EBS is mainly caused by autosomal dominant mutations in the genes KRT5 and KRT14 encoding keratins 5 (K5) and K14 (1, 2). Heterodimers of K5 and K14 form the keratin intermediate filaments (KIF) of the basal keratinocyte cytoskeleton. In EBS, skin fragility is thought to result from a defective assembly, growth and impaired stability of KIF (3, 4). Mutations, usually missense substitutions, cluster in distinct domains of keratins and the severity of clinical manifestations seems to correlate with the localisation of these mutations (5–7). In the most severe form, the Dowling-Meara subtype (EBS-DM; MIM 131760), mutations are almost all located in the highly conserved ends of the alpha-helical rod domain of K5 and K14, whereas in the mildest variants, the ‘localised EBS’ (previously Weber-Cockayne; MIM 131800) and the ‘EBS, other generalised’ (previously Köebner; MIM 131900), mutations affect less conserved sequences of the alpha-helical rod domain, such as the linker regions. Unusual EBS variants have also been recognised that are caused by recurrent mutations outside the rod domain, in the head and tail of K5 (1, 2). These mutations are believed to impair keratin filament elongation resulting in skin blistering and defective melanin granule aggregation. Mutation p.P25L in the K5 head domain is the most frequent substitution identified in the EBS with mottled pigmentation subtype (EBS-MP; MIM 131960) (8). Mutation c.1649delG, located in the tail domain of K5, causes the EBS with migratory circinate erythema (EBS-MCE; MIM 609352), a singular clinical phenotype characterised by arcuate migratory inflammatory blistering lesions that heal with brownish pigmentation (9, 10). The pathogenic mechanism through which the c.1649delG mutation would cause the peculiar lesions of EBS-MCE remains elusive. Here, we report on an EBS-MCE patient and document the presence of immunocompetent cells in an active lesion of the patient, who bears the c.1649delG variant in KRT5. Our results provide evidence for an involvement of the cellular immunity in the expression of the disorder.



The diagnosis of EBS was established during the neonatal period on the basis of the clinical features and confirmed by electron microscopy examination of a lesional skin biopsy. At the age of 3 years, a blood sample and a 4 mm punch biopsy from the edge of the migratory erythema were taken for genetic analysis and immunopathological studies, respectively. Biological samples were obtained after written informed consent of the parents. The study was conducted in compliance with the Declaration of Helsinki principles.

Mutation detection

Mutational analysis was performed by direct bidirectional sequencing of KRT5 exons, as described (11). A second PCR reaction and resequencing of the PCR product was performed to verify the mutation further. Total RNA was extracted from the patient skin biopsy using TRIzol reagents (Invitrogen, Carlsbad, CA, USA). The cDNA was synthesised using Superscript III RT (Invitrogen), and amplified with KRT5 intron-crossing primers that hybridise on exon 7 (5’-GGAGCTCATGAACACCAAGC) and exon 9 (5’-CTTGGAAGGCAGTGACTTGC).


The 4 mm punch biopsy, obtained at the age of 3 years from the active edge of a migrating lesion, was frozen at –80°C in OCT compound for immunofluorescence and immunohistochemical investigations and TUNEL staining. Immunofluorescence examination was performed on frozen 5-µm-thick sections using a three-step biotin-streptavidin-fluorescein procedure, as described (12). K5 was probed with monoclonal antibody (mAb) C50 (Laboratory Vision, Fremont, CA, 1:50 dilution).

For immunohistochemistry, 5-µm cryostat sections were fixed in 4% paraformaldehyde for 5 min and pre-treated for 15 min with PBS containing 0.3% hydrogen peroxide to block the endogenous peroxidase activity, and with protein block serum-free solution (DAKO, Glostrup, Denmark) to block the nonspecific binding of primary and secondary Abs. Slides were incubated for 1 h at room temperature in a humid atmosphere with 2.5 µg/ml of anti-CD1a (HI149, mIgG1), 5 µg/ml of anti-CD3 (HIT3a, IgG2a), 2.5 µg/ml of anti-CD4 (RPA-T4, mIgG1), 1.2 µg/ml of anti-CD8 (RPA-T8, mIgG1), 50 µg/ml of anti-CD69 (FN50, mIgG1) and 2.5 µg/ml of anti-CD83 (HB15e, mIgG1) mAbs (all from BD Bioscience,San Diego, CA), followed by biotinylated anti-mouse (DAKO) secondary Ab for 10 min. Streptavidin-HRP complex (DAKO) was added for 10 min to the samples, and 3-amino-9-ethyl-carbazole chromogen (Vector Laboratories, Burlingame, CA) was used as substrate. Every step was followed by extensive washing in PBS. Slides were counterstained with haematoxylin. As a negative control, primary Abs were omitted or replaced with irrelevant isotype-matched Abs.

TUNEL staining

In Situ Cell Death Detection Kit (Roche Applied Science, Penzberg, Germany) was used to identify DNA strand breaks by labelling free 3’-OH termini with fluorescein nucleotides in a template independent manner enzymatic reaction. Briefly, cryopreserved tissue sections were fixed with 4% paraformaldehyde for 5 min, incubated with 0.3% H2O2 for 15 min to inactivate endogenous peroxidase and permeabilised with 0.05% Triton for 5 min at room temperature. Afterwards, sections were incubated with terminal deoxynucleotidyl transferase enzyme and fluorescein nucleotides mixture for 60 min at 37ºC in a humidified chamber in the dark. After 3 washes in PBS, the fluorescein labels incorporated in nucleotide polymers were detected and quantified by fluorescence microscopy using an excitation wavelength of 488 nm and a detection of 515–565 nm.


Clinical features and laboratory findings

The patient was a female infant, first child of non-consanguineous parents. Since the 4th day of life, she presented trauma-induced blisters at first localised to hands and feet (Fig. 1a) that rapidly spread to the entire body surface. Blistering tendency markedly decreased after the first year of life, when lesions continued to form only on hands and feet and in other body areas exposed to trauma. By the age of 2 years the patient presented a generalised reticulated pigmentation affecting in particular the neck, axillary and inguinal folds, and limbs (Fig. 1b). In addition, circinate and migrating areas of erythema with multiple vesicles and small crusts at the advancing edge developed on the nape and shoulders (Fig. 1c). The erythematous lesions were extremely itchy and resolved with a patchy hyperpigmentation. The hair and nails were normal and mucosae were not affected. At subsequent follow-up the disappearance of the migratory erythematous lesions was noticed by the age of 4 years following sun exposure in summer. No erythematous migrating lesions were noticed for the subsequent 6 years (Fig. 1d). At the age of 10 years the patient suddenly developed a widespread eruption of itchy erythematous papules and small plaques, often annular-shaped or circinate, and with occasional tiny vesicles in the periphery (Fig. 1e, f). Additional findings include residual mottled pigmentation, in particular of axillary folds (Fig. 1e), mild onychodystrophy, mainly of the toenails, and rare trauma-induced foot blisters. Family history revealed that the mother had suffered from acral blistering during the first year of life followed by development of a reticulated pigmentation that disappeared by the age of 18 years. The mother and grandmother did not remember any migrating erythematous lesion. The mother still presents minimal onychodystrophy (Fig. S1; available from

Diagnosis of EBS-MCE in the index case was based on clinical examination and ultrastructural findings (not shown) in infancy. The latter showed cleavage within basal keratinocyte cytoplasm, in the absence of tonofilament clumping. Moreover, immunofluorescence analysis of the biopsy obtained at the age of 3 years revealed epidermal basal vacuolisation and an irregular labelling of basal keratinocytes for K5 (not shown). The clinical diagnosis was confirmed by KRT5 sequencing, which revealed the c.1649delG mutation at the heterozygous state in the DNA of the proband and her mother (Fig. 2). The expression of the mutant allele was confirmed by direct sequencing of a cDNA fragment, which show­ed frame shifted sequences across the mutation site. However, the nucleotide peaks corresponding to the mutant cDNA were lower than the wild-type cDNA. This finding indicates that the mutant mRNA is less abundant than the normal one, possibly due to a reduced stability (Fig. 2, bottom panel).


Fig. 1. Clinical features of the proband. Large blister of the ankle and toe erosion in the neonatal period (a). Diffuse mottled pigmentation, more evident on the upper limbs and neck (b), and migratory circinate erythematous vesicular and crusted lesions healing with patchy hyperpigmentation of the back (c) at the age of 3 years. Minimal residual reticulated pigmentation of the back, in the absence of any active lesion, at the age of 9.5 years (d). Annular erythematous small plaques and reticular hyperpigmentation of the axillary fold at the age of 10 years (e). Circinate and arcuate erythematous plaques and excoriated erythematous vesicles of the upper posterior thigh at the age of 10 years (f).


Fig. 2. Identification of the c.1649delG mutation in KRT5 and mRNA analysis. Sequence chromatogram of exon 9 amplified from the index case reveals a heterozygous nucleotide deletion within a run of 6 consecutive guanosine. The deletion of a guanosine is evident in the sequence of a subcloned allele. WT, wild-type sequence of KRT5 exon 9 from a normal control. Sequence analysis of a cDNA fragment encompassing the mutation site (bottom right panel). Note that the nucleotide peaks corresponding to the mutant cDNA are lower than the wild-type cDNA.

Immunohistochemical studies and TUNEL assay

The characteristics of the inflammatory infiltrate at the active border of an erythematous lesion were investigated by immunohistochemistry. We found that the majority of the inflammatory cells are CD3+ T lymphocytes with a slight prevalence of the CD4+ T-cell subset over the CD8+ subpopulation (Fig. 3a–c). Both CD4+ and CD8+ T cells were predominantly distributed in the upper dermis, but a substantial number of the infiltrating lymphocytes entered the epidermis in close contact with basal and suprabasal keratinocytes. Interestingly, the activation marker CD69 was expressed by more than 50% of infiltrating cells (Fig. 3d). The number of intraepidermal CD1a+ Langerhans’ cells was not increased (Fig. 3e). However, a high number of dendritic cells expressing the activation marker CD83 were present in the upper dermis (Fig. 3f). TUNEL staining revealed numerous apoptotic keratinocytes scattered throughout all epidermal layers in lesional skin (Fig. 3g).


The patient here reported confirms the tight phenotype-genotype correlation of the c.1649delG mutation in EBS-MCE (6, 9, 10). This deletion is predicted to cause a frame shift that alters the sequence of the last 41 amino acids and adds 35 residues to the tail of K5. However, mutation consequences at the mRNA and protein level have never been investigated in vivo. Our analysis on total RNA extracted from the patient skin biopsy showed a reduced amount of transcripts encoding for mutant K5 polypeptides. On the other hand, an in vitro study has demonstrated that the keratin filaments assembled from wild type K14 and mutant K5 lacking the last 41 residues are shortened and weakened in their viscoelastic properties when subjected to strain (13). These alterations can account for the skin fragility, but do not explain the inflammatory and migrating lesions. Moreover, the c.1649delG mutation has also been associated with EBS with mottled pigmentation (8, 14, 15). Our case shows features of both EBS-MCE, with typical lesions at first localised to the nape and back and later in a more widespread distribution, and EBS with mottled pigmentation, more evident on the limbs and large folds. In addition, the mother of the patient had a history compatible with EBS with mottled pigmentation alone. Therefore the c.1649delG mutation seems to result in the majority of patients in EBS with mottled pigmentation and in some cases to also cause MCE, implying that additional molecular modifiers are involved in MCE pathogenesis. The absence of EBS-MCE lesions in our patient during 6 years also suggests that environmental factors could be involved in triggering the inflammatory lesions in these patients.

The intervention of immune responses orchestrated by locally recruited dendritic cells that activate immune functions of skin T lymphocytes might partially explain the typical erythematous lesions occurring in MCE (16). Our results demonstrated the presence of a heavy leukocyte infiltrate at the edge of the advancing erythematous lesions. The presence of CD3+, CD4+ cells within the epidermis and, to a lesser extent, of CD8+ T-cell subsets in the upper dermis, was ascertained (Fig. 3a–c). Many infiltrating cells in physical contact with basal and suprabasal keratinocytes expressed the CD69 activation marker (Fig. 3d). Altogether, these findings support the hypothesis of a prominent role of the skin immune system in sustaining the clinical expression of the disease. Previous reports demonstrated the increase of the chemokines CCL2, CCL19 and CCL20 in the skin of keratin 5–/–, but not keratin 14–/–, mice (17, 18). CCL20 is a strong chemoattractant for T lymphocytes, Th17 cells in particular, and may thus be responsible for the accumulation of effector T lymphocytes in the skin, which could contribute to the keratinocyte damage. Indeed, the presence of apoptotic TUNEL-positive basal and suprabasal keratinocytes in lesional skin (Fig. 3g) suggests the role of perforin- and/or FAS/FasL-mediated T-cell cytotoxicity.


Fig. 3. Immunohistochemical staining of a biopsy from the advancing edge of the shoulder circinate lesion. Five-µm cryostat sections were fixed in paraformaldehyde and stained with the indicated antibodies as described in Methods. Results confirm the presence of numerous CD3+ T lymphocytes (a) belonging to both CD4 (b) and CD8 (c) subsets. Several infiltrating T lymphocytes express the CD69 activation marker (d). Although CD1a+ Langerhans’ cells are not increased in the epidermis (e), a high number of CD83+ dendritic cells can be detected in the dermis (f). The presence of the inflammatory infiltrate is paralleled by apoptotic TUNEL-positive basal and suprabasal keratinocytes (g). Magnification: ×200 (a–c, e–g), ×400 (d).

In line with the hypothesis of a prominent role of the immune system in the expression of the disease, MCE manifestations in our patient were cleared for several years after summer exposure to UV radiation, a known modulator of both innate and adaptive skin immune responses (Fig. 1).

Although at the advancing edge of the circinate inflammatory lesions we could not detect a significant increase of the number of intraepidermal Langerhans’ cells (Fig. 3e), as previously described in keratin 5–/– mouse and in patients affected with EBS by Roth et al. (17), our patient displayed a high number of dendritic cells in the upper dermis expressing the activation marker CD83 (Fig. 3f). This finding confirms the active recruitment and maturation of dendritic cells at the lesional site. Whether the inflammatory response depends on a specific immune response to neo-antigens released by damaged keratinocytes, or is the consequence of an unspecific response following the release of pro-inflammatory cytokines and chemokines in affected areas remains to be established.


This work was supported by grants from Italian Ministry of Health (RC11-1.2 and RC11-1.4).


1. Fine J-D, Eady RAJ, Bauer EA, Bauer JW, Bruckner-Tuderman L, Heagerty A, et al. The classification of inherited epidermolysis bullosa (EB): report of the third international consensus meeting on diagnosis and classification of EB. J Am Acad Dermatol 2008; 58: 931–950.

2. Sprecher E. Epidermolysis bullosa simplex. Dermatol Clin 2010; 28: 23–32.

3. Coulombe PA, Kerns ML, Fuchs E. Epidermolysis bullosa simplex: a paradigm for disorders of tissue fragility. J Clin Invest 2009; 119: 1784–1793.

4. Herrmann H, Strelkov SV, Burkhard P, Aebi U. Intermediate filaments: primary determinants of cell architecture and plasticity. J Clin Invest 2009; 119: 1772–1783.

5. Arin MJ, Grimberg G, Schumann H, De Almeida H Jr, Chang YR, Tadini G, et al. Identification of novel and known KRT5 and KRT14 mutations in 53 patients with epidermolysis bullosa simplex: correlation between genotype and phenotype. Br J Dermatol 2010; 162: 1365–1369.

6. Bolling MC, Lemmink HH, Jansen GH, Jonkman MF. Mutations in KRT5 and KRT14 cause epidermolysis bullosa simplex in 75% of the patients. Br J Dermatol 2011; 164: 637–644.

7. García M, Santiago JL, Terrón A, Hernández-Martín A, Vicente A, Fortuny C, et al. Two novel recessive mutations in KRT14 identified in a cohort of 21 Spanish families with epidermolysis bullosa simplex. Br J Dermatol 2011; 165: 683–692.

8. Echeverría-García B, Vicente A, Hernández A, Mascaró JM, Colmenero I, Terrón A, et al. Epidermolysis bullosa simplex with mottled pigmentation: a family report and review. Pediatr Dermatol 2012 May 29 [Epub ahead of print].

9. Gu LH, Kim SC, Ichiki Y, Park J, Nagai M, Kitajima Y. A usual frameshift and delayed termination codon mutation in keratin 5 causes a novel type of epidermolysis bullosa simplex with migratory circinate erythema. J Invest Dermatol 2003; 121: 482–485.

10. Nagao-Watanabe M, Fukao T, Matsui E, Kaneko H, Inoue R, Kawamoto N, et al. Identification of somatic and germline mosaicism for a keratin 5 mutation in epidermolysis bullosa simplex in a family of which the proband was previously regarded as a sporadic case. Clin Genet 2004; 66: 236–238.

11. Whittock NV, Eady RA, McGrath JA. Genomic organization and amplification of the human epidermal type II keratin genes K1 and K5. Biochem Biophys Res Commun 2000; 274: 149–152.

12. Kanitakis J, Zambruno G, Viac J, Tommaselli L, Thivolet J. Expression of an estrogen receptor-associated protein (p29) in epithelial tumors of the skin. J Cutan Pathol 1989; 16: 272–276.

13. Gu LH, Coulombe PA. Defining the properties of the nonhelical tail domain in type II keratin 5: insight from a bullous disease-causing mutation. Mol Biol Cell 2005; 16: 1427–1438.

14. Horiguchi Y, Sawamura D, Mori R, Nakamura H, Takahashi K, Shimizu H. Clinical heterogeneity of 1649delG mutation in the tail domain of keratin 5: a Japanese family with epidermolysis bullosa simplex with mottled pigmentation. J Invest Dermatol 2005; 125: 83–85.

15. Tang HY, Du WD, Cui Y, Fan X, Quan C, Fang QY, et al. One novel and two recurrent mutations in the keratin 5 gene identified in Chinese patients with epidermolysis bullosa simplex. Clin Exp Dermatol 2009; 34: e957–961.

16. Goldsmith LA. Migrating skin lesions: a genetic clue. J Invest Dermatol 2003; 121: vii–viii. Erratum in: J Invest Dermatol 2003; 121: 1565.

17. Roth W, Reuter U, Wohlenberg C, Bruckner-Tuderman L, Magin TM. Cytokines as genetic modifiers in K5-/- mice and in human epidermolysis bullosa simplex. Hum Mutat 2009; 30: 832–841.

18. Lu H, Chen J, Planko L, Zigrino P, Klein-Hitpass L, Magin TM. Induction of inflammatory cytokines by a keratin mutation and their repression by a small molecule in a mouse model for EBS. J Invest Dermatol 2007; 127: 2781–2789.