Filaggrin genotype in ichthyosis vulgaris predicts abnormalities in epidermal structure and function

Abstract

Although it is widely accepted that filaggrin (FLG) deficiency contributes to an abnormal barrier function in ichthyosis vulgaris and atopic dermatitis, the pathomechanism of how FLG deficiency provokes a barrier abnormality in humans is unknown. We report here that the presence of FLG mutations in Caucasians predicts dose-dependent alterations in epidermal permeability barrier function. Although FLG is an intracellular protein, the barrier abnormality occurred solely via a paracellular route in affected stratum corneum. Abnormal barrier function correlated with alterations in keratin filament organization (perinuclear retraction), impaired loading of lamellar body contents, followed by nonuniform extracellular distribution of secreted organelle contents, and abnormalities in lamellar bilayer architecture. In addition, we observed reductions in corneodesmosome density and tight junction protein expression. Thus, FLG deficiency provokes alterations in keratinocyte architecture that influence epidermal functions localizing to the extracellular matrix. These results clarify how FLG mutations impair epidermal permeability barrier function.

Figure 1

Cytoskeletal abnormalities in IV. As compared to control subjects (wt/wt) (A), intermediate filament retractions and reductions in F-type keratohyalin are visible in keratinocytes of FLG-deficient epidermis (B–D). Abnormalities shown are from a FLG compound heterozygote (B and C) and a homozygote (D) subject. Arrows and asterisks depict perinuclear keratin intermediate filament retraction. N, nucleus; SC, stratum corneum; SG, stratum granulosum. A and B: Toluidine blue staining. Scale bar = 20 μm. C and D: Osmium tetroxide postfixation. Scale bars = 0.2 μm.

Figure 2

Filaggrin genotype–dependent changes in epidermal function in IV. A: Hyperkeratosis assessed as the number of SC layers in coded, randomized, H&E-stained sections. wt, wild type. B: Basal barrier function, assessed as transepidermal water loss (TEWL) with an evaporimeter. C: Barrier recovery kinetics, expressed as percentage change in barrier function over time after sequential tape strippings (initial TEWL ≥threefold of basal level). D: SC hydration, assessed as electrical conductance with a corneometer. AU, arbitrary units. E: Surface pH, assessed with a flat-surface probe. F: Intercellular cohesion, assessed as application of successive tape strips followed by quantitation of total protein.

Figure 3

Altered corneocyte integrity in FLG-deficient IV. A: Higher percentages of fragile corneocytes (CEf) in heterozygotes and homozygotes compared to rigid corneocytes (CEr) in wild-type (wt) individuals. Corneocytes were treated with detergent (2% SDS) and heat (80°C), and assessed by phase microscopy. Original magnification, ×100. B: The time to disruption of sonicated corneocytes is significantly decreased in double-allele IV. C: Similar degrees of destruction after 90 seconds of ultrasound exposure are observed in FLG-deficient and wild-type subjects. D: Normal-appearing corneocyte envelopes (arrowheads) in IV. Ruthenium tetroxide postfixation. Scale bars = 0.25 μm.

Figure 4

Paracellular barrier defect in IV and FLG-deficient, organotypic keratinocyte cultures. A: Biopsies were incubated ex vivo with a water-soluble, low molecular weight tracer, colloidal lanthanum nitrate. Lanthanum tracer is excluded from both the corneocyte cytosol and the extracellular matrix of the SC in control subjects (wt/wt). B: Tracer breaches stratum granulosum (SG)–SC interface in both single-allele (not shown) and double-allele patients and crosses SC by a paracellular route. C–E: Filaggrin expression in organotypic skin cultures was reduced by >90% with both siRNAs before incubation with the lanthanum tracer. C: Stop of tracer perfusion at the SG–SC interface in keratinocytes transfected with a scrambled (scr)RNA. D and E: FLG-deficient cultures (two different siRNAs) display lanthanum movement into and across the SC via the paracellular pathway. Curved arrows (B–E) indicate (outward) direction of tracer movement. Small arrows (D and E) mark colloidal lanthanum nitrate in the extracellular space. Osmium tetroxide postfixation. Scale bars = 0.5 μm.

Figure 5

Abnormal extracellular lamellar bilayer organization and maturation in IV. A and C: Normal lamellar bilayers (arrows), corneodesmosomes (double-arrows), and corneodesmosome-derived lacunae (asterisks) in two different wild-type (wt) controls. B and D–E: Lamellar bilayers, corneodesmosomes, and corneodesmosome-derived lacunae in two different, double-allele IV subjects. B: Normal numbers of bilayers (arrows), but disruption of membrane arrays by nonlamellar domains (asterisks). D and E: Delayed maturation, ie, transformation of secreted lamellar body contents into bilayers above the SG–SC interface (asterisks). Note that corneodesmosome structure (double arrows) appears normal. Ruthenium tetroxide postfixation. Scale bars 0.25 μm (A–D); 0.2 μm (E).

Figure 6

Defective lamellar body (LB) contents in LB secretory system in two double-allele IV subjects (both are compound heterozygote for the same FLG mutations). A and B: Nonlamellar (vesicular) contents at the SG–SC interface (asterisks). C and D: Nonlamellar (vesicular) contents (asterisks) in single LBs, suggesting defective loading of organelle contents. SC, stratum corneum; SG, stratum granulosum. Osmium tetroxide postfixation. Scale bars = 0.2 μm.

Figure 7

Abnormalities in TJ protein expression in FLG-deficient epidermis shown by immunofluorescence staining for occludin (A, C, and E) and ZO-1 (B, D, and F). A and B: In control skin, a clear staining at the cell–cell borders of the uppermost living layers is observed for both occludin and ZO-1 (red). C and D: Reduced staining with increased cytoplasmic staining in heterozygous skin biopsies. E and F: Staining is almost completely lost in FLG-deficient skin. The nuclei are stained with DAPI (blue). Overlay of the epifluorescence pictures. Scale bar = 50 μm.

Figure 8

Basis for the barrier abnormality in IV. The two upper rows indicate structural alterations, whereas the lower rows indicate pathways for functional alterations.