An update on molecular aspects of the non-syndromic ichthyoses

Abstract

Abstract:  Ichthyosis includes a number of subtypes from congenital severe forms, such as harlequin ichthyosis (HI), to mild non-congenital forms, such as ichthyosis vulgaris. Recently, research into the pathomechanisms of ichthyoses has dramatically advanced and led to the identification of several causative genes and molecules underlying the genetic defects. In most types of ichthyosis, pathogenic mechanisms are associated with defects in skin barrier function. Three major components of the stratum corneum barrier are (i) intercellular lipid layers, (ii) cornified cell envelope and (iii) keratin-filaggrin degradation products. The causative molecules underlying ichthyosis subtypes include ABCA12, lipoxygenase-3, 12R-lipoxygenase, CYP4F2 homolog, ichthyin and steroid sulphatase and all these are thought to be related to the intercellular lipid layers. Transglutaminase 1 has a function in cornified cell envelope formation. Keratins 1, 10 and 2 are involved in the keratin network of suprabasal keratinocytes and filaggrin are essential for formation of keratohyalin granules. In fact, loss of ABCA12 function leads to a defective lipid barrier in the stratum corneum, resulting in the HI phenotype and ABCA12 is a known keratinocyte lipid transporter associated with lipid transport in lamellar granules. Filaggrin gene mutations in ichthyosis vulgaris cause keratohyalin granule deficiency. Information concerning genetic defects and ichthyotic disease pathomechanisms are beneficial to develop effective therapy and provide information for genetic counselling including prenatal diagnosis for families affected by ichthyotic disease.

Introduction

Ichthyoses, especially severe congenital ichthyoses, can be devastating to a patients’ quality of life in those seriously affected, even though other organs are uninvolved. Six major distinct clinical subtypes are known in hereditary non-syndromic ichthyoses, i.e. starting with the most severe form: harlequin ichthyosis (HI); lamellar ichthyosis (LI); non-bullous congenital ichthyosiform erythroderma (NBCIE); bullous congenital ichthyosiform erythroderma (BCIE); a mild and non-congenital form, X-linked ichthyosis (XLI); to the mildest form of ichthyosis vulgaris (IV) (Fig. 1). Ichthyosis bullosa of Siemens (IBS) is an additional subtype similar to BCIE. For a long time, the pathomechanisms and underlying genetic defects were unknown, although significant progress has recently been made in our understanding of the molecular basis of the human epidermal keratinization processes.

Figure 1.  Clinical features of ichthyosis. (a) Harlequin ichthyosis: thick, plate-like scales covered the entire body surface. (b) Lamellar ichthyosis: large, dark grey to black scales on the legs. (c) Ichthyosis vulgaris: fine, whitish scales on non-erythrodermic skin of the leg.

In 1978, the causative abnormality underlying XLI was clarified as a steroid sulphatase deficiency caused by genetic defects in the steroid sulphatase gene (STS) (1,2). In 1992, mutations in the keratin 1 (KRT1) and keratin 10 genes (KRT10) were detected as a cause of BCIE (3–5). As transglutaminase 1 gene (TGM1) mutations were identified as the cause in LI in 1995 (6,7), mutations in several other genes have also been identified in severe autosomal recessive congenital ichthyoses (ARCI).

In 2005, loss of function mutations in the ABCA12 gene were reported to underlie HI, the most severe type of ichthyosis (8,9). In 2006, null mutations in the gene coding filaggrin (FLG) were detected as the causative defects leading to IV (10).

To date, in total the number of genes that have been identified and demonstrated to cause ichthyosis in human patients has reached eleven and these are shown in Table 1 (8,11–18), i.e. FLG (10), KRT1, KRT10 (3–5), KRT2 (19–21), TGM1 (6,7), ABCA12 (8,16,22), two lipoxygenase genes, ALOXE3 and ALOX12B (23), ichthyin (17) and FLJ39501 (24). Besides these ichthyosis variants with known causative genes, there are three monogenetic variants of the ichthyoses reported, for which linkage has been established but no underlying genetic defect has been described yet concerning to the three loci, 3p21 (14), 12p11.2-q13 (18) and 19p13.2-p13.1 (13).

Table 1.   Major three components of stratum corneum barrier and causative molecules for ichthyoses

Stratum corneum barrier components	Molecule	Gene (locus)	Mode of inheritance	Type of mutations	Phenotype
AD, autosomal dominant; AR, autosomal recessive; ASD, autosomal semidominant; X-LR, X-linked recessive; BCIE, bullous congenital ichthyosiform erythroderma; HI, harlequin ichthyosis; IBS, ichthyosis bullosa of Siemens; IV, ichthyosis vulgaris; LI, lamellar ichthyosis; NBCIE, non-bullous congenital ichthyosiform erythroderma; XLI, X-linked ichthyosis.
1. Intercellular lipid layers	ABCA12	ABCA12 (2q34)	AR	Truncation/deletion (rarely missense)	HI
	ABCA12	ABCA12 (2q34)	AR	Missense	LI or NBCIE
	lipoxygenase-3	ALOXE3 (17p13.1)	AR	Missense/truncation	LI or NBCIE
	12R-lipoxygenase	ALOX12B (17p13.1)	AR	Missense/truncation	LI or NBCIE
	Cytochrome P450 (CYP4F2 homolog)	FLJ39501 (19P12)	AR	Missense/truncation	LI
	Ichthyin	ichthyin (5q33)	AR	Missense/truncation	NBCIE or LI
	Steroid sulphatase	STS (Xp22.32)	X-LR	Mostly large deletion	XLI
2. Cornified cell envelope	Transglutaminase 1	TGM1 (14q11.2)	AR	Missense/truncation/ deletion/insertion	LI or NBCIE
3. Keratin network and  keratohyalin granules	Keratin 1	KRT1 (12q12-q13)	AD	Missense	BCIE
	Keratin 10	KRT10 (17q21)	AD (rarely AR)	Missense (rarely nonsense)	BCIE
	Keratin 2	KRT2 (12q11-q13)	AD	Missense	IBS
	Filaggrin (profilaggrin)	FLG (1q21.3)	ASD	Truncation	IV

From the information on the pathogenesis of the ichthyoses mentioned above, a majority of cases show primary abnormal phenomena that are associated with barrier function in the stratum corneum (Fig. 2). The skin barrier of the stratum corneum has three major components and mechanisms of pathogenesis of ichthysosis are thought to be associated with at least one of them (Fig. 3). One is the intercellular lipid layers in the stratum corneum. Another is the cornified cell envelope of the stratum corneum cells. The other one is keratin-filaggrin degradation products filling the cytoplasm of cornified cells in the stratum corneum. In this review, the pathogeneses of various ichthyoses are described in association with descriptions of specific defects in each of the essential components comprising the epidermal skin barrier, highlighting a few crucial diseases and mechanisms that stand as representative examples for the still more complex pictures.

Figure 2.  Skin barrier of cornified cells and the intercellular lipid layers can reduce transepidermal water loss (TEWL) (green arrows) and can limit access of exogenous allergens or irritants (blue arrows).

Figure 3.  Three major components of the skin barrier in the stratum corneum.

Major three components of stratum corneum barrier and causative molecules for ichthyoses

Intercellular lipid layers in the stratum corneum

It is of little doubt that defective formation of stratum corneum intercellular lipid layers is the due to abnormal keratinocyte lipid metabolism, transport, and/or secretion as one of the major pathogenetic mechanisms underlying congenital ichthyosis. Several critical molecules causing ichthyosis are thought to be involved in the formation of the stratum corneum intercellular lipid layers and these will be described in the following sections.

ABCA12

Formation of the intercellular lipid layers is essential for epidermal barrier function and defective formation of the lipid layers is thought to result in serious loss of barrier function (Fig. 4) and to extensive hyperkeratosis (25,26).

Figure 4.  Cornified cell envelope and intercellular lipid layers: normal and defective formation. (a) Model of formation of the normal intercellular lipid layers and cornified cell envelope in the stratum corneum. (b) ABCA12 dysfunction leads to defective intercellular lipid layers in HI. (c) TGase 1 deficiency results in a malformed cornified cell envelope and subsequent defective intercellular lipid layers in LI.

Formation of the intercellular lipid layers is a highly complicated series of processes that include transport of lipids into the lamellar granules and multi-step metabolism of this lipid content within lamellar granules. ABCA12 has been highlighted, because ABCA12 was recognized as a key molecule in keratinocytes lipid transport (8,27).

In 2003, ABCA12 was reported as the causative gene in type 2 LI (OMIM #601277) (16). Type 2 LI is a subtype of LI which links to 2q33-35. Clinically, no distinct characteristic features are known for this type of LI (11,12,16). Among the severe ARCI diseases, HI is the most devastating congenital ichthyosis and affected newborns show large, thick, plate-like scales over the whole body with severe ectropion, eclabium and flattened ears (28). In 2005, we revealed that ABCA12 is a keratinocyte lipid transporter and demonstrated that ABCA12 mutations lead to the HI phenotype (8). Another group independently reported that ABCA12 mutations underlie HI by linkage analysis (9). In addition, ABCA12 mutations were reported to underlie certain NBCIE cases (29). ABCA12 is a member of a large superfamily of the ATP-binding cassette (ABC) transporters, which bind and hydrolyze ATP to transport various molecules across a limiting membrane or into a vesicle (30). All ABCA subfamily members are thought to be lipid transporters (31). ABCA12 is a keratinocyte transmembrane lipid transporter protein associated with lipid transport in lamellar granules to the apical surface of granular layer keratinocytes (8) (Fig. 4).

Ultrastructurally, lamellar granule abnormalities are apparent in HI patient epidermis (32–35). Several morphological abnormalities have been reported, for example abnormal lamellar granules in the granular layer keratinocytes and a lack of extracellular lipid lamellae in the stratum corneum, reflect defective lipid transport by lamellar granules and the malformation of intercellular lipid layers in the stratum corneum in HI (32–36).

In addition, cultured epidermal keratinocytes from an HI patient carrying ABCA12 mutations demonstrated defective glucosylceramide transport and this phenotype was recoverable by in vitro ABCA12 corrective gene transfer (8). From these findings, we have shed light on the pathomechanisms of HI with the underlying cause being ABCA12 mutations that leading to a loss of ABCA12 function. Mutations in the lipid transporter protein, ABCA12, cause defective lipid accumulation into lamellar granules, which then become expelled from the apical surface of keratinocytes (8,37), resulting in malformation of the intercellular lipid layers of the stratum corneum (8). The fact that ABCA3 (a member of the same protein superfamily as ABCA12) functions in pulmonary surfactant lipid secretion again via the production of similar lamellar-type granules within lung alveolar type II cells (38,39) further supports this concept.

Several genotype/phenotype correlations with ABCA12 mutations have now come to light, including combinations of missense mutations resulting in only one amino acid alteration that underlie the LI phenotype (16). However, most mutations in HI are truncation or deletion mutations which lead to more severe changes, such as loss of function of ABCA12 peptide affecting important nucleotide-binding fold domains and/or transmembrane domains. In HI, thus far at least one mutation on each allele must be a truncation or deletion mutation within a conserved region that seriously affects ABCA12 function (8,9,40–45). Only a couple of cases showing a NBCIE phenotype were reported to have ABCA12 mutations as the causative genetic defect (29). Further accumulation of data on ABCA12 mutations and their effects on proteins together with specific mutation sites is needed to better elucidate genotype/phenotype correlations and help predict HI patient prognosis once novel combinations of ABCA12 mutations have been identified.

Recently, we transplanted keratinocytes from patients with HI and succeeded in reconstituting HI skin lesions in immunodeficient mice (37). The reconstituted lesions showed similar changes to those of patients with HI. This reconstitution system of HI skin lesions could be a powerful tool to analyze ABCA12 function and to evaluate various treatment for HI (37).

Lipoxygenase-3 and 12(R)-lipoxygenase

In 2002, mutations in two lipoxygenase genes, ALOXE3 and ALOX12B, coding lipoxygenase-3 and 12(R)-lipoxygenase, respectively, were reported to underlie ARCI (23). Lipoxygenase-3 and 12(R)-lipoxygenase are non-haeme iron-containing dioxygenases expressed in the epidermis, and their exact functions are unknown (46–48). They may be associated with lipid metabolism of the lamellar granule contents and/or intercellular lipid layers in the epidermis. 12(R)-lipoxygenase knockout mice demonstrated a severe impairment of skin barrier function (49). The loss of barrier function was associated with perturbance of the assembly/extrusion of lamellar granules. Cornified cell envelope from skin of 12(R)-lipoxygenase deficient mice showed increased fragility (49). Lipid analysis demonstrated a disordered composition of ceramides, especially a decrease of ester-bound ceramide species (49). From these findings, the 12(R)-lipoxygenase-lipoxygenase-3 pathway was thought to play a key role in the process of epidermal barrier formation by affecting lipid metabolism (49).

Cytochrome P450, family 4, subfamily F, polypeptide 2 homolog

Mutations in FLJ39501 were identified in lamellar ichthyosis type 3 (MIM 604777) as causative genetic defects (24). FLJ39501 encodes a cytochrome P450, family 4, subfamily F, polypeptide 2 homolog of the leukotriene B4-omega-hydroxylase (CYP4F2) and is thought to catalyze the 20-hydroxylation of trioxilin A3 form the 12(R)-lipoxygenase pathway. Further oxidation of this substrate by the fatty alcohol oxidoreductase (FAO) or nicotinamide-adenine dinucleotide oxidoreductase enzyme complex, in which one component of the enzyme system, ALDH3A2, is mutated in Sjögren–Larsson syndrome, would lead to 20-carboxy-(R)–trioxilin A3. This compound is suspected of being involved in skin hydration and would be an essential product missing in many forms of ARCI. Observations in Sjögren–Larsson syndrome harbouring fatty aldehyde dehydrogenase (FALDH) gene (ALDH3A2) mutations demonstrated defective lamellar granule formation as a link to the putative pathogenetic mechanism producing ichthyotic lesions in these patients’ skin (50).

Ichthyin

Ichthyin defects were also reported to underlie certain cases of LI or NBCIE phenotype (17). Ichthyin is a protein with several transmembrane domains, which belongs to a new family of proteins with an as yet unknown function. Ichthyin-like proteins are localized in the plasma membrane, and share homologies to both transporters and G-protein coupled receptors (17). Ichthyin was suggested to be a membrane receptor for certain ligands (trioxilins A3 and B3) from the hepoxilin pathway (17). However, the exact mechanisms of how ichthyin mutations can cause an ichthyotic phenotype remain to be clarified.

Steroid sulphatase

Genetic defects in the steroid sulphatase gene (STS) was reported to underlie XLI (1,2). The hyperkeratosis and scaling in XLI is associated with abnormal accumulation of cholesterol sulphate in the stratum corneum (51). Steroid sulphatase is concentrated in lamellar granules and secreted into the intercellular space of stratum corneum, along with other lamellar granule-derived lipid hydrolases (52). In the intercellular space of stratum corneum, steroid sulphatase degrades cholesterol sulphate, generating some cholesterol for the barrier. In addition, the progressive decline in cholesterol sulphate permits corneodesmosome degradation leading to normal desquamation (52). Thus, two molecular pathways contribute to disease pathogenesis in XLI. Steroid sulphatase deficiency leads both to malformation of the intercellular lipid layers leading to barrier defects, and delay of corneodesmosome degradation, resulting in corneocyte retension (52). Furthermore, increased Ca²⁺ in the intercellular space of the stratum corneum in XLI was reported to contribute to corneocyte retention, by increasing corneodesmosomes and interlamellar cohesion (52).

Cornified cell envelope in lamellar ichthyosis

Another important event for the skin barrier function during epidermal keratinocyte terminal differentiation is the formation of a 15-nm-thick layer of protein on the inner surface of the cell membrane, called the cornified cell envelope (CCE) (Fig. 4). Transglutaminase 1 which is essential in proper CCE formation is known to be a causative molecule for ichthyosis.

Transglutaminase 1

Cornified cell envelope is assembled by accumulation of several precursor proteins including involucrin, small proline-rich proteins and loricrin (53). It is known that these precursors are cross-linked together by the formation of N^ε-(γ-glutamyl) lysine isodipeptide bonds catalysed by the action of transglutaminases (TGases) expressed in the epidermis (54,55). TGase 1, the major subtype of the three TGases expressed in the epidermis (56,57), is a membrane-associated TGase of about 92 kDa. Transglutaminases in the epidermis are thought to be responsible at least in part, for the assembly of cornified cell envelope precursor proteins that go to form the cornified cell envelope (55).

As the identification of TGase 1 gene (TGM1) mutations in a number of families with LI (MIM #242300) in 1995 (6,7), additional TGM1 mutations have been reported in LI families. Furthermore, TGM1 mutations were reported to underlie NBCIE phenotype (58,59).

As unique clinical entities of congenital ichthyosis caused by TGM1 mutations, which were elucidated at a molecular level, self-healing collodion baby and bathing suit ichthyosis should be mentioned here. Self-healing collodion baby heals with no or only very mild ichthyosis. TGM1 mutations p.Gly278Arg and p.Asp490Gly were found in self-healing collodion babies with markedly diminished epidermal transglutaminase 1 activity (60). In these patients, molecular modelling and biochemical assays suggested significantly reduced enzyme activity in p.Gly278Arg and a chelation of water molecules in p.Asp490Gly that locks the mutated enzyme in an inactive trans conformation under elevated hydrostatic pressure in utero (60). After birth, the water molecules are removed and the enzyme is thought to isomerize back to a partially active cis form, explaining the dramatic improvement in the phenotype. Bathing suit ichthyosis is a congenital ichthyosis caused by TGM1 mutations, characterized by pronounced scaling on the bathing suit areas but sparing of the extremities and the central face (61,62). In situ transglutaminase assay in the epidermis of bathing suit ichthyosis patients revealed a marked decrease of enzyme activity when the temperature was increased from 25 to 37°C. Bathing suit ichthyosis is now thought to be a temperature-sensitive phenotype (61).

Interestingly, CCE formation has been shown to be important for building the intact intercellular lipid layers. Transglutaminase 1 is an enzyme that was reported to crosslink hydroxyceramide to the cornified cell envelope (63,64). Cornified cell envelope formation is an essential scaffold upon which normal intercellular lipid layer formation in the stratum corneum can then take place (65). Thus, mutations in the TGM1 secondarily cause defects in the intercellular lipid layers in the stratum corneum, leading to defective barrier function of the stratum corneum and to the ichthyotic phenotype seen in LI patients (65) and in transglutaminase 1 knockout mice (66). It still remains to be determined how much a defective cornified cell envelope alone contributes to the barrier abnormality in ichthyoses.

Keratin network and keratohyalin granules

A normal keratin filament network is an important basic structure for keratohyalin granule formation and for maintaining the integrity and dimensions of the cornified cell cytoplasm that will later hold the keratin-filaggrin degradation products. In this context, the keratin-network and keratohyalin granules are both essential for normal skin barrier formation. Mutations in either keratins or profilaggrin/filaggrin are now known to cause forms of ichthyosis.

Keratin 1 and keratin 10

Bullous congenital ichthyosiform erythroderma is caused by mutations in either the keratin 1 or keratin 10 gene (3–5,67). Most of the causative mutations are missense mutations that reside within the beginning or at the end of the rod domain segments of the K1 or K10 peptide, called either the helix initiation or helix termination motif.

In all members of the keratin protein family, the helix boundary motifs are highly conserved regions of approximately 20 amino acids located at the beginning and the end of the central coiled-coil rod domain, which have been implicated in molecular overlapping interactions as part of the formation of 10 nm intermediate filaments from dimers comprising both type I acidic and a type II basic-neutral keratins (68). Single amino acid point mutation alterations in these essential, helix boundary motifs frequently cause a significant disease phenotype in the majority of the keratin diseases [Human Intermediate Filament Mutation Database (http://www.interfil.org)].

Bullous congenital ichthyosiform erythroderma is a severe congenital ichthyosis, which exhibits widespread blisters and erosions on a background of erythrodermic skin from birth. BCIE generally shows autosomal dominant inheritance. After the perinatal period, blister formation ceases and generalized hyperkeratosis becomes apparent. Histologically, predominant vacuolization of the cells is observed in the middle and upper spinous layers and granular layers. The vacuolated keratinocytes have large and irregularly shaped cytoplasmic granules. Ultrastructurally, irregularly shaped, abnormal, clumped keratin filaments, mainly comprising K1 and K10, are seen in the keratinocytes of the upper spinous layer to the granular layer (69,70).

Recent genotype–phenotype correlations in BCIE showed that palmoplantar keratoderma exists in patients with keratin 1 gene mutations but not in patients with keratin 10 gene mutations (71,72).

Only one K10 knockout family was reported and these patients show a recessive inheritance trait (73). In this recessive family, the unaffected, mutation heterozygous carriers indicated that an unaffected K10 peptide from the one normal allele alone is sufficient for keratin network formation.

Keratin 2

IBS (MIM #146800) is an autosomal dominant hereditary keratinization disorder which shows similar, but slightly milder clinical and histopathological findings than BCIE (MIM #113800) (74). In 1994, keratin 2 (K2) mutations were identified as causative in IBS patients (19–21). Occasionally, cases with IBS can be difficult to clinically differentiate from BCIE and molecular genetic studies are essential for a more definite diagnosis (75). In the human epidermis, K1 and K10 expression occurs in the suprabasal layers, replacing keratin 5 and keratin 14 as cells differentiate. K2 is expressed somewhat later than K1 and K10 in keratinocyte differentiation as keratinocytes approach the granular layer. Thus, consistent with the restricted K2 expression sites, in IBS, clumped keratin filaments were restricted to the cytoplasm of granular layer cells and the uppermost spinous layer cells, leading to granular degeneration only in the uppermost spinous and granular layers of the patient’s epidermis.

The granular degeneration restricted to the uppermost spinous and granular layers results in the milder clinical manifestations and the presence of superficial denuded areas over hyperkeratotic skin (the mauserung phenomenon) which are characteristic to IBS.

Filaggrin

As previously described, degradation products of keratohyalin granules subsequently occupy the cytoplasm of keratinized cells in the stratum corneum and play important roles in the skin barrier function. Keratohyalin granules in the granular layer of the epidermis are predominantly composed of large >400 kDa profilaggrin polyproteins (76–78). Upon terminal differentiation of keratinocytes, profilaggrin is dephosphorylated and cleaved into 10–12 essentially identical 37-kDa filaggrin peptides. The liberated filaggrin subsequently and highly efficiently aggregates the keratin filament cytoskeleton (76), causing the collapse of the granular cells into classic flattened squame shape. The collapsed cytoskeleton is cross-linked by TGases to link it to the cornified cell envelope. Another function has been suggested whereby degradation products of filaggrin contribute to moisture retention in the cornified layers (79). Thus, filaggrin, a major component of keratohyalin granules, is a key epidermal protein and essential for the formation of a normal, intact and protective skin barrier. Loss or reduction of filaggrin expression correlates with excessively dry skin and impaired barrier function, which variously manifests as IV.

Ichthyosis vulgaris (OMIM #146700) is a relatively common genetic keratinization disorder, clinically characterized by scaling, especially on the flexor limbs, and with palmoplantar hyperlinearity (80). Histologically, a decrease in the size and number or, complete absence of keratohyalin granules in the epidermis is typically characteristic of IV (80).

In 2006, two null mutations in the FLG gene were identified in IV patients from Scottish, Irish and European-American populations (10). Until now, a total of 15 FLG mutations have been reported and seven of them are prevalent in at least each population studied thus far (81). The prevalent filaggrin mutations are loss-of-function alleles leading to complete or almost complete loss of filaggrin expression (81,82).

It has been known that patients with atopic dermatitis (AD) often show ichthyotic skin and that patients with IV often develop AD. FLG mutations have recently attracted great attention as a key predisposing factor for AD (83). Disruption of the normal epidermal barrier due to filaggrin gene mutations allows allergens to penetrate into skin and leads to their exposure to antigen presenting cells, resulting in allergic sensitization. In some AD patients, the development of asthma and allergic rhinitis will follow, that is, the so-called ‘atopic march’ (84). Recently, FLG null mutations have been reported to predispose allergic phenotypes to become more likely to be involved in the atopic march (82,85–87). These facts are a milestone in molecular genetic research in atopic disorders and the next step will be translating all this knowledge on the disturbance of barrier function into more rational and more effective treatments for AD (88).

Prenatal diagnosis of severe congenital ichthyoses

The quality of life of patients with the more severe congenital ichthyoses is often seriously affected and the parents’ request for prenatal diagnosis is not easily ignored. From the 1990s, the genes causing severe congenital ichthyosis subtypes have been clarified one by one. Due to the recent advances in our understanding of the causative genetic defects for severe congenital ichthyosis, it has now become possible to make DNA-based prenatal diagnosis for several congenital ichthyosis by chorionic villus or amniotic fluid sampling procedures in the earlier stages of pregnancy with a lower risk to foetal health and with a reduced burden on the mothers.

In the case of HI, before the causative genes were identified, prenatal diagnosis had been performed by foetal skin biopsy and electron microscopic observation during the later stages of pregnancies at 19–23 weeks estimated gestational age for more than 20 years (34,89–91). When a foetus was diagnosed as affected, it was a major problem to perform interruption of the pregnancy in that late stage of pregnancy. As previously described earlier in the article, in 2005, the underlying gene causing HI was identified as ABCA12 (8). Since this discovery, it has become possible to make HI DNA-based prenatal diagnosis by chorionic villus or amniotic fluid sampling in the earlier stages of pregnancy (44), as in the case of other severe genetic disorders.

Prenatal diagnosis of LI had been possible by ultrastructural observation of foetal skin samples, although prenatal diagnosis of LI was somewhat high risk among all the keratinization diseases because LI patients can show regional, individual and familial variability in their disease manifestation (92). In LI families with TGM1 mutations, successful prenatal diagnosis and prenatal exclusion of LI by TGM1 mutation analysis has already been reported (93,94). Theoretically, prenatal diagnosis by mutation analysis in lipoxygenase-3, 12(R)-lipoxygenase and ABCA12, etc. is available in LI families with previously identified mutations on a case-by-case basis and is dependent on the severity of gene mutations.

Successful prenatal diagnosis for BCIE by foetal skin biopsy was reported in 1980s (95) and nowadays prenatal diagnosis by direct gene sequencing is possible for BCIE (96,97).

For severe congenital ichthyosis, as is the case with other genetic conditions, perhaps even earlier prenatal diagnosis by completely noninvasive analysis of DNA from foetal cells in maternal circulation (98), and preimplantation genetic diagnosis will be available in the future (99,100).

Novel therapeutic approaches to ichthyoses

The skin is the most easily accessible organ for direct topical and injectable gene transfer in a minimally invasive and simple manner (101).

In case of HI, phenotype recovery by corrective gene transfer was demonstrated in HI patient derived keratinocytes harbouring ABCA12 mutations (8). Recently, we succeeded in reconstitution of human HI skin lesions using HI patient’s cultured keratinocytes harbouring ABCA12 mutations transplanted on the back of immunodeficient mice (37). This reconstituted HI skin model is a powerful tool to test further gene therapy experiments. Alternatively, ex vivo gene therapy has been reported for LI, where normal gene expression of TGM1 has been restored and the phenotype correction was observed in engrafted lesional skin in vivo on the back of immunodeficient mice (102). In the case of Sjögren–Larsson syndrome, successful gene transfer of functional FALDH using adeno-associated virus-2 vectors to FALDH deficient human keratinocytes was reported and, furthermore, three-dimensional culture of the corrected patient’s keratinocytes demonstrated FALDH expression (103). However, in any type of ichthyosis, significant further development for any practical trials of gene therapy in humans has yet to be achieved.

In our animal experiments, hair follicles were reconstituted from hair follicle stem cell cultures carrying transgenes (104). The transgene expression was observed in all skin epithelial compartments including the epidermis for at least 6 months. This hair follicle stem cell targeted gene transfer and reconstitution system provides reliable gene-function analysis and gene therapy (104). These advances will stimulate the future research for in vivo gene therapy for ichthyosis patients.

Besides the efforts to induce gene therapy, other forms of therapy including enzyme substitution or up-regulation of enzyme activity have been attempted. For example, a hypolipidemic drug bezafibrate is a pan-agonist of all peroxisome proliferator-activated receptor (PPAR) isoforms and was reported to induce FALDH activity in fibroblasts of Sjögren–Larsson syndrome patients with ALDH3A2 gene mutations (105). Thus, bezafibrate treatment could be effective in Sjögren–Larsson syndrome patients with some residual FALDH enzyme activity (105).

Acknowledgements

I would like to thank Professor James R. McMillan for his proofreading and comments during the preparation of the manuscript.

This work was supported in part by Grant-in-Aid from the Ministry of Education, Science, Sports, and Culture of Japan to M. Akiyama (Kiban B 18390310) and to H. Shimizu (Kiban A 17209038).