An abnormal synthesis or metabolism of the LG lipid contents was previously suspected as being a possible pathogenetic mechanism underlying HI (11
). Here, we demonstrate that a severe ABCA12 deficiency causes defective lipid transport via LG in keratinizing epidermal cells, resulting in the HI phenotype.
The ABC transporter superfamily is one of the largest gene families, encoding a highly conserved group of proteins involved in energy-dependent active transport of a variety of substrates across membranes, including ions, amino acids, peptides, carbohydrates, and lipids (3
). ABC transporters have nucleotide-binding folds located in the cytoplasm and utilize energy from ATP to transport substrates across the cell membrane (3
). ABC genes are widely dispersed throughout the eukaryotic genome and are highly conserved between species (6
). The ABCA subfamily, of which the ABCA12
gene is a member, comprises 12 full transporter proteins and 1 pseudogene (ABCA11). The ABCA subclass has received considerable attention (14
) because mutations in these genes have been implicated in several human genetic diseases (15
The ABCA1 protein is mutated in the following recessive disorders involving cholesterol and phospholipid transport: Tangier disease (MIM 205400), familial hypoalphalipoproteinemia (MIM 604091), and premature atherosclerosis, depending on the site of the mutations in the protein (17
). The ABCA4 protein is mutated in Stargardt disease (MIM 248200) as well as in some forms of autosomal recessive retinitis pigmentosa (MIM 601718) and in the majority of cases of autosomal recessive cone-rod dystrophy (MIM 604116), depending on the mutation site or the combination of mutation types (15
). Heterozygous mutations in ABCA4
have also been implicated in some cases of macular degeneration (MIM 153800) (16
). In a relatively mild type of congenital ichthyosis (lamellar ichthyosis type 2), 5 ABCA12
mutations were reported in 9 families, and all 5 of these mutations were missense mutations that resulted in only 1 amino acid alteration (8
). In the present series of HI patients, no ABCA12
missense mutations were identified, and most of the defects led to severe truncation of the ABCA12 peptide, affecting important nucleotide-binding fold domains and/or transmembrane domains. The other, nontruncation mutations in HI were deletion mutations affecting highly conserved ABCA12
sequences (Figure D). Thus, it is thought that only truncation or conserved region deletion mutations that seriously affect the function of the ABCA12 transporter protein lead to the HI phenotype. This is in contrast with diseases caused by mutations in other members of the ABCA family. Of ABCA4
mutations causing Stargardt disease, 80% were missense, and many of these occurred in conserved domains of ABCA4 (21
). Of ABCA1
mutations resulting in Tangier disease, 60% were missense, located within the conserved domains of ABCA1 (22
). In this context, ABCA12
mutations underlying HI are unique in that these mutations are restricted to truncation or deletion mutations.
Patient 1 was homozygous for the splice acceptor site mutation IVS23-2A→G. This splice-site mutation was shown to lead to comparable amounts of 2 predicted transcripts, an inframe deletion of 3 amino acids, and a transcript with a 170-nucleotide deletion (frameshift) (Figure , A–C). Thus, it is still possible that patient 1 expresses some mutated ABCA12 protein. Indeed, expression of a small amount of ABCA12 protein, although mutated, was detected in the granular cells of the patient’s epidermis and cultured keratinocytes by immunofluorescent staining (Figures C and J). This might have some relevance to the fact that patient 1 survived infancy and is still alive.
ABCA1 and ABCA4 are suspected transmembrane transporters for intracellular cholesterol/phospholipids (23
) and protonated N
), respectively. ABCA2
, and ABCA7
mRNA levels have been reported to be upregulated by sustained cholesterol influx mediated by modified low-density lipoprotein (27
), suggesting that ABCA transporters are involved in transmembrane transport of endogenous lipids (23
). Keratinocyte LGs are known lipid-transporting granules, and LG contents are secreted into the intercellular space, forming an intercellular lipid layer between the granular layer cells and keratinized cells in the stratum corneum, which is essential for the proper barrier function of human skin. Our results indicate that ABCA12 functions in the transport of endogenous lipid across the keratinocyte cell membrane into the stratum corneum intercellular space via LGs. Immunohistological and immunoelectron microscopic observations have indicated that ABCA3 is involved in lipid secretion of pulmonary surfactant in human lung alveolar type II cells (29
). ABCA3 and ABCA12 are very closely related in the ABCA subfamily phylogenetic tree (30
). It is interesting that the functions of both ABCA3 and ABCA12 are involved in alveolar surfactant and stratum corneum lipid secretion, respectively. This suggests that these 2 transporter systems are evolutionary adaptations to aid the respiratory system and the integument in a dry environment, developed as vertebrates left the aquatic environment and began terrestrial lives. HI skin that harbors serious defects in ABCA12 highlights the important role(s) of epidermal lipid transport in adapting human skin to a terrestrial, dry environment.
HI is one of the most severe of all genodermatoses and has a very poor prognosis. Thus, parents’ request for prenatal diagnosis is to be taken seriously and with care. However, to our knowledge, the causative gene was not identified until now. For the last 20 years, prenatal diagnoses have only been performed by electron microscopic observation of fetal skin biopsy at a late stage of pregnancy (19–23 weeks estimated gestational age) (31
). In this report, we have identified the causative gene in HI, which now makes possible DNA-based prenatal diagnosis using chorionic villus or amniotic fluid sampling at an earlier stage of pregnancy, with a lower procedural risk and a reduced burden on prospective mothers, similar to that of screening for other severe genetic disorders, including keratinization disorders (34
). Furthermore, we have performed corrective gene transfer in HI keratinocytes and succeeded in obtaining phenotypic rescue of a patient’s cultured keratinocytes. These data provide significant clues that establish a strategic approach to HI gene therapy treatment. We believe that the genetic information presented in this study will be highly beneficial to our understanding of HI pathogenesis and in optimizing HI patient diagnosis, genetic counseling, care, and treatment.