Fig 1.
Skin abnormalities in smsk mice.
(a) E18.5 smsk embryos display a shiny, tight and thickened skin phenotype with limb contractures. (b) Smsk E18.5 embryos have epidermal barrier defects as demonstrated by a toluidine dye penetration assay. (c) Histology of mutant E18.5 embryonic skin demonstrates severe hyperkeratosis with more than 16 layers of corneocytes in the SC and disorganized architecture in other epidermal layers. Bar for a and b = 2.5 mm. Bar for c = 25 μm. Epidermal layers: SB = stratum basale; SS = stratum spinosum; SG = stratum granulosum; SC = stratum corneum.
Fig 2.
A mutation in Abca12 is associated with the smsk phenotype.
(a). The recessive mutation responsible for the skin phenotype of smsk was mapped by linkage analysis to a region containing the Abca12 genomic locus. Genomic DNA sequencing showed that the wild type (WT) 5’ splice donor site “GT” in the 29th intron of Abca12 on the C57/B6 background was mutated to “GG” in the phenotypic mice. Non-phenotypic carrier mice contained both “GT” and “GG” alleles. (b) Alignment of WT and mutant Abca12 cDNA sequences showed that the splice site mutation resulted in a skipping of exon 29 in the mutant Abca12 mRNA and would produce an in-frame deletion of amino acids encoded by exon 29 in the mutant protein. (c) The cDNA fragments of Abca12 encompassing exon 28 to 31 were amplified by RT-PCR from skin RNA of WT (764bp) and smsk mutant (HI, 545bp) mice. The bands were resolved by agarose gel electrophoresis. (d) In silico translation analysis predicted that the smsk mutation results in a truncated protein with an in-frame fusion between amino acid sequences encoded by exons 28 (green) and 30 (red). The normal sequence encoded by the splicing of exon 28 to exon 29 is shown in black. (e) Alternative splicing skips exon 29, leading to an in-frame deletion of 73 amino acids from the first ATP-binding domain of ABCA12.
Fig 3.
Defects in lipid accumulation in the stratum corneum of Abca12smsk/smsk mutant mice.
(a) Immunofluorescence staining for glucosylceramide/ceramide (GlcCer/Cer) showed localization throughout the suprabasal layers of WT epidermis, whereas reduced levels were detected in the SC (thin dashed lines) of E18.5 Abca12smsk/smsk epidermis. Bar = 25 μm. (b) Transmission electron microscopy (TEM) shows the disappearance of corneodesmosomes (CDs) (white arrows) above the SG-SC interface in WT mice, whereas CDs are retained in smsk SC. Bars = 200 nm and 500 nm respectively. (c) TEM pictures show the presence of normal intercellular lipid lamellae (arrow) at the junctions between SG and SC layers in the WT epidermis but not in the mutant epidermis. Bars = 200 nm. (d) Ultrastructural analysis shows that lamellar bodies (LBs) in WT epidermis were loaded with lipid lamellae and fused with the surface of granular cells (arrow). LBs in mutant epidermis had no lamellar cargo, but fusion with the granular cell membrane appeared normal. Bars = 200 nm.
Fig 4.
Abca12smsk/smsk displays defects in terminal differentiation.
(a). Cell proliferation rates were similar in both WT and smsk embryonic epidermis as determined by in utero incorporation of 5-bromo-2'-deoxyuridine (BrdU). Sections were stained with an anti-BrdU antibody (green) and counterstained with KERATIN 14 (KRT14) (red). Note that no difference in the expression pattern of KRT14 was observed in either WT or mutant epidermis. Bars = 25 μm. (b) Keratinocytes in Abca12smsk/smsk mutant skin undergo abnormal terminal differentiation as demonstrated by scattered pattern of KRT1 (red) detection in the SC. Bars = 50 μm. (c) KRT16 (red) was ectopically expressed in the interfollicular suprabasal keratinocytes of Abca12smsk/smsk epidermis, whereas it was undetecteable in the WT epidermis. Furthermore, the retention of nuclei (parakeratosis) in the SC is observable in smsk mutant skin. Bar = 25 μm.
Fig 5.
Enhanced adhesion of cornified envelopes in the Abca12smsk/smsk epidermis.
E18.5 skin pieces from both WT and Abca12smsk/smsk embryos were boiled in cornified envelope (CE) extraction buffer. (a) WT skin pieces dissociated after boiling for 10 minutes, resulting in the dissolution of skin into individual CEs (Left panels). The mutant skins were never completely dissolved, even after extended hours of boiling. No individual CEs from the mutant skin were evident (Right panels). Bars = 5 mm and 20 μm, respectively. (b-f) Skin sections of both WT and Abca12smsk/smsk embryos showed (b) expression of DESMOPLAKIN (DSP), a component of both desmosomes and CDs, persisting throughout the thickened SC of the smsk epidermis; (c) a minor decrease in DESMOCOLLIN (DSC3) staining in the smsk SC; (d) no change in DESMOGLEIN (DSC) 1&2 staining in mutant skin versus WT; (e) reduced SC localization of CORNEODESMOSIN (CDSN) in mutant skin; (f) and INVOLUCRIN (IVL) staining throughout the SC of mutant mice but reduced levels in lower layers compared to WT skin. (g-h) Immunofluorescence detection of KLK5 and -7 revealed their presence in all layers of the WT epidermis with prominent expression observed at the junction between granular and cornified layers. In contrast, KLK staining was reduced in the smsk SC. Bars = 25 μm.
Fig 6.
Inhibition of glucosylceramide production mimics defects in KALLIKREIN secretion from cultured Abca12smsk/smsk keratinocytes.
(a) Secretion of KLKs into the culture media from cultured keratinocytes were measured via ELISA. Basal levels of KLK5 and -7 were readily detected in the culture media of keratinocytes grown under low Ca++ conditions, and this increased after differentiation induced by elevated Ca++ levels (p<0.01 for both KLK5 and -7). Pre-treating WT keratinocytes with d,l-threo-1-phenyl-2-decanoylamino-3-morpholino-1-propanolhydro-chloride (PDMP) significantly reduced secreted KLK5, -7 levels (p<0.05). Note that levels of KLKs in the media of Abca12smsk/smsk cells remained unaffected by Ca++ or PDMP treatment. Both KLK and protease activity in smsk keratinocytes were lower than WT (**, p<0.05). (b) Protease activity was measured in the conditioned media from WT or Abca12smsk/smsk keratinocytes. Similar to secreted KLK levels, protease activity in WT cells was elevated after Ca++ differentiation (p<0.05) and suppressed by PDMP pretreatment. In contrast, protease activity in the cultured media remained unaffected in Abca12smsk/smsk keratinocytes either after differentiation or PDMP treatment. Sample size n = 3 was used for each condition tested.
Fig 7.
Transplant of Abca12smsk/smsk skin grafts.
(a) Back skins of E18.5 WT and Abca12smsk/smsk embryos were transplanted onto nude mice. Bar = 2 mm. (b) Histological presentation of H&E stained sections of the grafts from WT and smsk transplanted skin. Note the loss of normal epidermal architecture at the end of 3 wks of transplantation in the smsk grafts. Bar = 50 μm. (c) Smsk E18.5 skin grafts were treated daily with a cream containing either the recombinant KLKs or the digestion buffer (Control). Shedding of the top layers of the hyperkeratotic skin grafts was observed only in grafts treated topically with KLK cream. Images were taken after 7 days of treatment. Bar = 2 mm. (d) Representative H&E images are shown of control and KLK cream treated smsk mutant skin grafts. Bar = 10 μm.