Kindler syndrome (KS) was reported first in 1954 as an unusual inherited skin condition in a 14-y-old girl with trauma-induced blister formation, nail dystrophy, progressive poikiloderma, hyperkeratosis of the palms and soles, webbing between toes and fingers, and sensitivity to sunlight (Kindler, 1954). A number of similar patients have subsequently been identified who have helped define KS as a distinct autosomal recessive disorder (OMIM 173650).
Many of the clinical features of blistering and scarring in children with KS resemble those seen in the inherited skin disorder dystrophic epidermolysis bullosa, and the underlying skin pathology often involves morphologically and structurally similar disruption of skin adhesion at the dermal–epidermal junction (Fine et al, 2000). The type VII collagen gene, COL7A1, however, which is mutated in dystrophic epidermolysis bullosa, has been excluded as a candidate gene for KS (Shimizu et al, 1997;Yasukawa et al, 2002). Moreover, transmission electron microscopy in KS may reveal additional pathologic findings including blister formation within basal keratinocytes, somewhat similar to cases of epidermolysis bullosa simplex, a disorder known to result from mutations in the genes encoding the intermediate filament proteins, keratins 5 or 14 (Irvine and McLean, 1999). Alternatively, there may be features of reduplication of the lamina densa, or variable and inconsistent changes, including multiple planes of cleavage as well as alterations in elastic fibers (Hovnanian et al, 1989).
Similar to patients with severe forms of epidermolysis bullosa, there is an increased risk of mucocutaneous malignancy in KS (Couwenberg et al, 1998;Lotem et al, 2001;Mallipeddi, 2002), but there are also further clinical differences, notably in the patients' response to ultraviolet irradiation. In KS, patients have an abnormal early erythema response to sun exposure as well as signs of progressive poikiloderma. Thus, for several years KS has been thought to represent an unusual, but distinct, variant of epidermolysis bullosa or an inherited poikiloderma syndrome.
Using genome-wide linkage analysis on DNA obtained from consanguineous families, two groups have recently mapped the KS gene to 20p12.3 and identified 10 different pathogenic homozygous loss-of-function mutations in a novel gene (Jobard et al, 2003;Siegel et al, 2003). This gene, KIND1, encodes a 677 amino acid protein, kindlin-1. It is well conserved throughout evolution, with closely related homologs in Drosophila and Onchorhynchus, as well as Caenorhabditis elegans, suggesting that it has an important function across species (Rogalski et al, 2000;Schaller, 2000;Siegel et al, 2003). Humans possess three differentially expressed kindlin genes (encoding kindlin-1, kindlin-2, and kindlin-3), of which kindlin-1 is the major kindlin expressed by basal epidermal keratinocytes (Siegel et al, 2003).
In this study, we searched for KIND1 mutations in 16 individuals with Kindler syndrome from 13 families to extend the mutation database and to identify any recurrent mutations in this genodermatosis that might have relevance to optimizing mutation detection strategies.
Results
Clinical Figure 1
Affected individuals 1–8 were from six Pakistani families residing in the UK. Cases 2 and 3 were siblings, as were cases 4 and 5, but none of the six families was known to be related to the others. Consanguinity was present within the individual families, however. Case 1 lived in southern England but the other families were spread across middle and northern England. The subjects' ages ranged from 3 mo to 30 y, there were six males and two females, and all had similar features of trauma-induced blistering, especially on the hands and feet, since infancy. Photosensitivity also began in infancy and consisted of erythema within minutes of sun exposure. Poikiloderma was often not apparent until 8–10 y and several cases were incorrectly diagnosed as dystrophic epidermolysis bullosa in early childhood. Most blistering ceased during teenage years. Thereafter, poikiloderma with prominent hyperpigmentation and hypopigmentation was the main clinical feature. Other features in adulthood included ectropion (cases 4 and 5) and dysphagia (case 5). In all children under the age of 7 y, the key investigation in establishing the diagnosis was transmission electron microscopy. Ultrastructurally, there was reduplication of the lamina densa, as well as variable focal areas of collagen lysis in the superficial dermis and occasional microclefts within the lamina lucida (data not shown).
Kindler syndrome case 9 (female; 7 y) was from the Al Mudhabi region of Oman, whereas Omani case 10 (female; 19 y) was from the isolated island of Museira. Geographically and socially, their families were not known to be related, although both contained consanguineous marriages. The clinical features were similar to the British Pakistani KS cases. In addition, case 10 also had dysphagia with esophageal stenosis. There was also a history of a sibling of case 10 dying in early infancy because of severe skin fragility and secondary infection.
Cases 11 (male; 23 y), 12 (female; 10 y), and 13 (female; 44 y) were unrelated white Caucasians from different parts of England. Although the clinical history of initial trauma-induced blistering and sun sensitivity was similar to the KS individuals above, the degree of poikiloderma was more subtle, reflecting the paler skin type (I–II; cf. types IV–V in subjects 1–10). The clinical features in case 11 closely resembled a mild form of dystrophic epidermolysis bullosa or possible non-Herlitz junctional epidermolysis bullosa, but again, reduplication of the lamina densa seen on transmission electron microscopy provided the clue to the correct diagnosis of KS.
Affected Italian individuals 14 and 15 were brothers (34 and 36 y) from the rural south of the country, whereas case 16 (male; 46 y) was a seemingly unrelated individual from urban central Italy. All had trauma-induced blistering from birth until adolescence. Features of poikiloderma were subtle and more closely resembled postinflammatory pigmentation. In case 16, immunohistochemical analysis of epidermal basement membrane disclosed reduced immunostaining for type XVII collagen, but no pathogenic mutation has been detected in subsequent sequencing of the COL17A1 gene (GZ, unpublished data). At the age of 43 y, case 16 developed two squamous cell carcinomas on the upper lip and dorsum of the hand, both of which were excised and there have been no signs of any recurrence in 3 y follow-up.
Delineation of new pathogenic KIND1 mutations in Kindler syndrome
Sequencing of genomic DNA from affected individuals in the 13 families with KS disclosed pathogenic mutations in each case, as shown in Figure 2. In all affected individuals from the six Pakistani families (n=8), genomic DNA sequencing revealed a homozygous insertion of cytosine at position 676 in exon 5 of KIND1, denoted 676insC Figure 2a (GenBank no. AY137240; nucleotide numbers refer to cDNA with methionine initiation codon ATG as 1–2–3, etc.). In the two Omani families, sequencing of affected individuals' DNA disclosed a homozygous guanine to adenine change at nucleotide 1848 in exon 14 resulting in replacement of a tryptophan residue (TGG) by a stop codon (TGA), denoted W616X Figure 2b. In the three UK Caucasian patients, sequencing showed a heterozygous guanine to thymine change at nucleotide 910 in exon 7 that changes a glutamic acid residue (GAA) to a stop codon (TAA), denoted E304X Figure 2c. Compound heterozygosity for a further nonsense or frameshift mutation was shown in each case. These mutations comprised a thymine to adenine substitution at nucleotide 905 in exon 7, which changes leucine (TTA) to a stop codon (TAA), L302X Figure 2c; a single nucleotide deletion in exon 10, 1161delA; and a single nucleotide deletion in exon 15, 1909delA. In the two Italian families, three affected individuals were homozygous for an acceptor splice site mutation 958–1G > A in intron 7 Figure 2d. These new mutations, and the previously documented mutations in other patients with KS, are illustrated in Figure 3.
Figure 2.
Sequencing reveals recurrent nonsense and frameshift mutations in KIND1. (a) Pakistani patients 1–8 show a homozygous insertion of a cytosine at nucleotide 676 in exon 5, denoted 676insC. (b) Omani patients 9 and 10 have a homozygous guanine to adenine substitution at nucleotide 1848 in exon 14 changing a tryptophan residue (TGG) to a stop codon (TGA), denoted W616X. (c) UK Caucasian patients 11–13 are all heterozygous for a guanine to thymine change at nucleotide 910 resulting in glutamic acid (GAA) being converted to a stop codon (TAA) in exon 7, denoted E304X. Each of these cases has a further heterozygous nonsense mutation on the other allele. Illustrated here in patient 11 is a thymine to adenine substitution at nucleotide 905 changing leucine (TTA) to a stop codon (TAA), L302X, depicted by the double arrowhead. (d) Italian patients 14–16 are homozygous for an acceptor splice site mutation, with a substitution of guanine by adenine in intron 7, denoted 958–1G > A. The start of exon 8 is indicated by an open bracket.
Full figure and legend (68K)Figure 3.
Schematic diagram showing the positions of all pathogenic mutations found in KIND1 and their positions within the gene. Mutations shown in bold are recurrent mutations, of which those identified in this study are underlined. Mutations in italics are also potential hotspot mutations. Exons (numbered 1–15) are to scale; introns in between are not to scale. Hatched areas denote start of 5' and 3' untranslated regions.
Full figure and legend (13K)Haplotype analysis reveals evidence for recurrent ancestral KIND1 mutations
Sequencing of genomic DNA in control and affected individuals disclosed that the KIND1 gene contains a number of highly polymorphic single nucleotide polymorphisms. Nine intragenic single nucleotide polymorphisms were identified. These comprised 1–29T/G, 114T/C, 151+20C/T, 152–4G/A, 479T/C, 532+8T/C, 532+34C/T, 533–17 A/C, and 722T/C, using the standard numbering system described above. Haplotype analysis of these polymorphisms in the affected individuals showed that the mutations W616X, E304X, and 958–1G > A occurred on similar genetic backgrounds for each mutation, consistent with propagation of common ancestral alleles in the Omani, UK Caucasian, and Italian gene pools, respectively Table I. Likewise, in five of the six Pakistani families, the mutation 676insC was shown on the same allelic background. In one case (patient 1), however, the mutated allele displayed an alternative haplotype with six out of nine intragenic polymorphisms being different.
Table I - Ancestral KIND1 alleles associated with recurrent pathogenic mutations underlying Kindler syndrome.
Full table Immunofluorescence microscopy reveals loss of expression of kindlin-1 in KS skin
In control skin, immunofluorescence labeling with antikindlin-1 antibody showed bright, pan-epidermal staining, especially within the basal keratinocyte cytoplasm and at the dermal-epidermal junction Figure 4a. No specific staining was seen in the dermis. In contrast, all labeling in patient skin was markedly reduced and in some cases absent Figure 4b, c, d, e.
Figure 4.
Immunofluorescence labeling with antikindlin-1 C-terminal antibody. (a) Control skin shows strong epidermal labeling within the basal keratinocytes and along the dermal–epidermal junction. (b) Patient 6, homozygous for the mutation 676insC, shows weak basal keratinocyte staining and broken staining along the dermal–epidermal junction. (c) Patient 11, compound heterozygous for E304X/L302X, shows weak staining throughout the epidermis. (d) Patient 10, homozygous for W616X, shows absence of staining. (e) Patient 16, homozygous for 958–1G > A, shows weak staining throughout the epidermis. (f) Negative control skin labeling. Bar: 40 
m.
Discussion
In this study, we have identified seven new pathogenic mutations in KIND1 in patients with KS, bringing the total found thus far to 17, as summarized in Figure 3. Previously, two recurrent mutations, R271X and R288X, were identified (Siegel et al, 2003). The KIND1 mutation R271X was demonstrated in Panamanian, Caucasian American, and Omani subjects, whereas the mutation R288X was found in UK Caucasian and Turkish individuals with KS. Both of these mutations were shown on different respective genetic backgrounds consistent with hypermutable CpG dinucleotides (Cooper and Krawczak, 1993) producing potential hotspots for loss-of-function mutations in kindlin-1. By contrast, our study has delineated four new KIND1 mutations, 676insC, W616X, E304X, and 958–1G>A, occurring on the same genetic background in the Pakistani, Omani, UK Caucasian, and Italian gene pools, respectively. These mutations indicate propagation of an ancestral allele through different generations of these populations and are important starting points in the investigation into the molecular pathology of further cases of KS from the same ethnic background. Separate from this study, we have also identified the presence of the homozygous frameshift mutation 1714delA in an Italian individual with KS of North African Arabic origin (unpublished data). Interestingly, this mutation has already been reported in another subject of similar genetic background (Jobard et al, 2003), raising the possibility of a further common mutated ancestral allele in this population.
Also noteworthy from our study is the Pakistani mutation 676insC, which appears to be both a hotspot and a founder mutation. This interpretation is supported by one individual (patient 1) having a different genetic background in KIND1 to the other seven affected subjects. The mutation 676insC occurs in a sequence of seven consecutive cytosines and may therefore have arisen in patient 1 through slipped mispairing (Cooper and Krawczak, 1993), compared to the propagation of a mutant allele in patients 2–8 inclusive.
The immunofluorescence microscopy studies in patients with any of these four new recurrent mutations in KIND1 (as well as those reported bySiegel et al, 2003) demonstrate the usefulness of the carboxy-terminal antikindlin-1 antibody as a potential diagnostic probe in KS. The marked reduction (or complete absence) of immunostaining in KS skin with this antibody appears to be a consistent and reliable finding. From a clinical perspective, it is clear that in nearly all the patients we studied there was a considerable delay in establishing a correct diagnosis of KS and refuting differential diagnoses of dystrophic epidermolysis bullosa or a poikilodermatous genodermatosis. Thus, the new antibody now allows for a rapid means of diagnosing KS.
Nevertheless, it should be noted that immunostaining was not completely absent in all KS subjects (see Figure 4). Notably, some weak labeling was noted in patients with the homozygous mutations 676insC and 958–1G > A, as well as the individual who was a compound heterozygote for the mutations E304X/L302X. These mutations are sited in exon 5, intron 7, and exon 7, respectively. By contrast, antikindlin-1 antibody labeling in skin homozygous for exon 14 mutation W616X was completely absent. The relative differences may relate to putative alternative splicing of KIND1 (Siegel et al, 2003). Indeed, preliminary assessment of potential KIND1 splice variants suggests that truncated isoforms spanning exons 2–8 and 9–15 may be expressed in normal keratinocytes (GHSA, unpublished data). The latter transcript would not be affected by mutations in exons 2–8 but could still react with the antibody made to the carboxy-terminal peptide in exon 15. This might explain the faint positive immunostaining in skin from cases with mutations upstream from exon 8, although other factors such as the consequences of the different mutations on RNA processing and transcript stability may influence the actual immunolabeling observed. From a clinicopathologic perspective, no major phenotypic differences were noted in patients with faint kindlin-1 immunostaining compared to those with a complete absence of antibody labeling. Despite the need for comprehensive characterization of normal splice variants in KIND1, the antibody remains a useful tool in establishing a diagnosis of KS. That said, one of the observations in the initial study characterizing the KIND1 gene was that some families with purported KS did not map to the KS locus on 20p12.3 (Siegel et al, 2003). This suggests that there may be a disorder that is clinically similar to but genetically different from KS. We have not yet had the opportunity to examine skin from such patients, however, to test the specificity of the new antikindlin-1 antibody. Similarly, one of the families reported by another research group showed evidence of genetic linkage to the KIND1 locus but the authors were unable to identify a mutation in this case (Jobard et al, 2003). In cases such as these, immunostaining for kindlin-1 in patient skin would be very useful to establish a diagnosis of KS in the absence of mutation data.
The predicted protein structure of kindlin-1 reveals several domains of interest, in particular a bipartite FERM (band 4.1, ezrin, radixin, and moesin) domain separated by a pleckstrin homology domain, indicative of roles in plasma membrane adhesion structures and a possible link to signal transduction (Chishti et al, 1998;Maffucci and Falasca, 2001;Lemmon et al, 2002;Siegel et al, 2003). Kindlin-1 is involved in linking the actin cytoskeleton to the extracellular matrix (ECM) and has been demonstrated in vivo to locate at the ends of filamentous actin fibers within the cell, at the point where they insert into focal contacts, and thus to the ECM (Siegel et al, 2003). This makes KS the first inherited skin fragility disorder to be associated with a defect in actin cytoskeleton ECM linkage, unlike subtypes of epidermolysis bullosa, which are associated with defects in keratin intermediate filament-ECM linkage. This observation may shed light on the unusual morphology seen on transmission electron microscopy in KS, which displays features different to other inherited skin fragility disorders, including extensive basement membrane reduplication (Shimizu et al, 1997;Siegel et al, 2003). The actual mechanisms leading to basement membrane dysregulation remain to be elucidated, however. Likewise, the actin cytoskeleton–focal contact link to the clinical features of photosensitivity, poikiloderma, and squamous cell carcinoma is still a mystery to be explored. Identification of the specific interactions of kindlin-1 with other molecules by the yeast two-hybrid system or similar strategies should shed light on these unresolved issues.
In essence, our study expands the mutation database of KIND1 mutations in KS, reports the utility of a new antikindlin-1 antibody in the rapid diagnosis of KS, and provides useful data in optimizing mutation screening in this genodermatosis.
Materials and Methods
PCR amplification of genomic DNA and mutation detection
Following approval by the ethics committee of St. Thomas Hospital, London, and informed consent, DNA was extracted from peripheral blood samples taken from affected individuals, their parents, and clinically normal siblings (where possible) using a standard cold water lysis method (Sambrook et al, 1989). PCR amplification of the KIND1 gene was performed as described elsewhere (Siegel et al, 2003). For PCR amplification, 500 ng of genomic DNA was used as template in an amplification buffer containing 1 M of each primer, 1.5 mM of MgCl2, 50 M of each nucleotide, and 2.5 U Taq polymerase (PE Biosystems, Warrington, UK) in a total volume of 50 L in a PE Biosystems 9700 thermal cycler. The amplification conditions were 95°C for 5 min, followed by 35 cycles of 95°C for 45 s, annealing temperature (seeSiegel et al, 2003) for 45 s, and 72°C for 45 s. Aliquots (5 L) of the PCR products were analyzed by 2% agarose gel electrophoresis. PCR products were purified using QIAquick PCR Purification Kit (Qiagen, Crawley, UK) and sequenced directly using Big Dye labeling in an ABI 310 genetic analyzer (PE Biosystems). Potential mutations were confirmed by restriction endonuclease digestion or bidirectional sequencing and assessed in up to 200 control chromosomes, selected to match the patients' ethnic backgrounds. The control samples were selected from 3000 archival DNA samples kept within the Genetic Skin Disease Group at St John's Institute of Dermatology, London, UK.
Identification of common nonpathogenic polymorphisms in KIND1
To search for common sequence variants in KIND1 that would be useful for haplotype analysis, genomic DNA from 40 control individuals of mixed ethnic backgrounds was amplified using the PCR primers and methods detailed above (and inSiegel et al, 2003) and screened using heteroduplex analysis, as described elsewhere (Ganguly et al, 1993). PCR products displaying bandshifts were then investigated further by direct nucleotide sequencing.
Immunofluorescence microscopy
Skin biopsies were available from four patients (patients 6, 10, 11, and 16). Skin was washed in phosphate-buffered saline (PBS) for 30 min before being embedded in OCT compound (TAAB, UK) and snap frozen in isopentane cooled by liquid nitrogen. Five micron cryosections were fixed in 1:1 acetone and methanol at -20°C for 20 min, then rehydrated in PBS for 2
15 min, and incubated in 0.1% Triton X-100 in PBS prior to blocking with 10% normal goat serum in PBS for 20 min. Sections were then incubated in the same well with primary kindlin-1 rabbit polyclonal antibody for 1 h at 37°C. Details of the kindlin-1 antibody (made against the carboxy-terminal 15 amino acids) have been published elsewhere (Siegel et al, 2003). After washing, sections were incubated with goat antirabbit Alexa Fluor 488 conjugate (Cambridge BioSciences, Cambridge, UK) for 1 h. A negative control was treated with the same method but replacing antikindlin-1 antibody with 10% normal goat serum. The samples were thoroughly washed in PBS and subsequently mounted on coverslips. Microscopy was performed using a Nikon Optiphot Microscope with Kodak Microscopy Document System 290.
References
| 1. | Chishti AH, Kim AC & Marfatia SM. et al The FERM domain: A unique module involved in the linkage of cytoplasmic proteins to the membrane. Trends Biochem Sci (1998) 23: 281–282. | Article | PubMed | ISI | ChemPort | |
| 2. | Cooper DN & Krawczak M. Human Gene Mutation. (1993) Oxford: BIOS Scientific Publishers Ltd. |
| 3. | Couwenberg SM, Vuzevski VD, Heule F & Siersema PD. Kindler syndrome. Ned Tijdschr Geneeskd (1998) 142: 1579–1580. | PubMed | ChemPort | |
| 4. | Fine JD, Eady RA & Bauer EA. et al Revised classification system for inherited epidermolysis bullosa. Report of the Second International Consensus Meeting on diagnosis and classification of epidermolysis bullosa. J Am Acad Dermatol (2000) 42: 1051–1066. | Article | PubMed | ISI | ChemPort | |
| 5. | Ganguly A, Rock MJ & Prockop DJ. Conformation-sensitive gel electrophoresis for rapid detection of single-base differences in double-stranded PCR products and DNA fragments: Evidence for solvent-induced bends in DNA heteroduplexes. Proc Natl Acad Sci USA (1993) 90: 10325–10329. | PubMed | ChemPort | |
| 6. | Hovnanian A, Blanchet-Bardon C & de Prost Y. Poikiloderma of Theresa Kindler: Report of a case with ultrastructural study, and review of the literature. Pediatr Dermatol (1989) 6: 82–90. | PubMed | ISI | ChemPort | |
| 7. | Irvine AD & McLean WHI. Human keratin diseases: The increasing spectrum of disease and subtlety of the phenotype–genotype correlation. Br J Dermatol (1999) 140: 815–828. | Article | PubMed | ISI | ChemPort | |
| 8. | Jobard F, Bouadjar B & Caux F. et al Identification of mutations in a new gene encoding a FERM family protein with a pleckstrin homology domain in Kindler syndrome. Hum Mol Genet (2003) 12: 925–935. | Article | PubMed | ISI | ChemPort | |
| 9. | Kindler T. Congenital poikiloderma with traumatic bulla formation and progressive cutaneous atrophy. Br J Dermatol (1954) 66: 104–111. | PubMed | ISI | ChemPort | |
| 10. | Lemmon MA, Ferguson KM & Abrams CS. Plekstrin homology domains and the cytoskeleton. FEBS (2002) 513: 71–76. | Article | ISI | ChemPort | |
| 11. | Lotem M, Raben M, Zeltser R, Landau M, Sela M, Wygoda M & Tochner ZA. Kindler syndrome complicated by squamous cell carcinoma of the hard palate: Successful treatment with high-dose radiation therapy and granulocyte-macrophage colony-stimulating factor. Br J Dermatol (2001) 144: 1284–1286. | Article | PubMed | ISI | ChemPort | |
| 12. | Maffucci T & Falasca M. Specificity in pleckstrin homology (PH) domain membrane targeting: A role for a phosphoinositide-protein cooperation mechanism. FEBS (2001) 506: 173–179. | Article | ISI | ChemPort | |
| 13. | Mallipeddi R. Epidermolysis bullosa and skin cancer. Clin Exp Dermatol (2002) 27: 616–623. | Article | PubMed | ISI | ChemPort | |
| 14. | Rogalski TM, Mullen GP, Gilbert MM, Williams BD & Moerman DG. The UNC-112 gene in Caenorhabditis elegans encodes a novel component of cell–matrix adhesion structures required for integrin localization in the muscle cell membrane. J Cell Biol (2000) 150: 253–264. | Article | PubMed | ISI | ChemPort | |
| 15. | Sambrook J, Fritsch EF & Maniatis T. Molecular Cloning: A Laboratory Manual. (1989) Cold Spring Harbour, NY: Cold Spring Harbour Laboratory Press. |
| 16. | Schaller MD. UNC112: A new regulator of cell–extracellular matrix adhesions? J Cell Biol (2000) 150: F9–F11. | Article | PubMed | ISI | ChemPort | |
| 17. | Shimizu H, Sato M & Ban M. et al Immunohistochemical, ultrastructural, and molecular features of Kindler syndrome distinguish it from dystrophic epidermolysis bullosa. Arch Dermatol (1997) 133: 1111–1117. | Article | PubMed | ISI | ChemPort | |
| 18. | Siegel DH, Ashton GH & Penagos HG. et al Loss of kindlin-1, a human homolog of the Caenorhabditis elegans actin–extracellular-matrix linker protein UNC-112, causes Kindler syndrome. Am J Hum Genet (2003) 73: 174–187. | Article | PubMed | ISI | ChemPort | |
| 19. | Yasukawa K, Sato-Matsumura KC, McMillan J, Tsuchiya K & Shimizu H. Exclusion of COL7A1 mutation in Kindler syndrome. J Am Acad Dermatol (2002) 46: 447–450. | Article | PubMed | ISI | |
Acknowledgments
We are grateful to the patients for their participation in this study, and to Dr E. Epstein Jr and Dr D. Siegel for initial collaborative studies on the KIND1 gene. The current work was supported by grants from Action Research, the Dystrophic Epidermolysis Bullosa Research Association (DEBRA, UK), the Barbara Ward Children's Foundation, and the British Skin Foundation. WHIM and FJDS are funded by a Wellcome Trust Senior Fellowship (to WHIM).
