HMCN1 variants aggravate epidermolysis bullosa simplex phenotype

Abstract

Epidermolysis bullosa simplex (EBS) refers to a heterogeneous group of inherited skin disorders characterized by blister formation within the basal cell layer. The disease is characterized by marked variations in phenotype severity, suggesting co-inheritance of genetic modifiers. We identified three deleterious variants in HMCN1 that co-segregated with a more severe phenotype in a group of 20 individuals with EBS caused by mutations in KRT14, encoding keratin 14 (K14). HMCN1 codes for hemicentin-1. Protein modeling, molecular dynamics simulations, and functional experiments showed that all three HMCN1 variants disrupt protein stability. Hemicentin-1 was found to be expressed in human skin above the BMZ. Using yeast-2-hybrid, co-immunoprecipitation, and proximity ligation assays, we found that hemicentin-1 binds K14. Three-dimensional skin equivalents grown from hemicentin-1-deficient cells were found to spontaneously develop subepidermal blisters, and HMCN1 downregulation was found to reduce keratin intermediate filament formation. In conclusion, hemicentin-1 binds K14 and contributes to BMZ stability, which explains the fact that deleterious HMCN1 variants co-segregate with a more severe phenotype in KRT14-associated EBS.

Figure 1.

Pedigrees, clinical features, and variant analysis. (A) Pedigrees of the four families. Black symbols denote affected individuals, red symbols denote more severely affected individuals. Arrows point to probands in each family. Genotypes are indicated for those individuals who consented to participate in the study. Below are electropherograms obtained through direct sequencing of HMCN1 that revealed a heterozygous G>A transition (arrow) at position c.11905 of the cDNA sequence in individual III–1, family 1, a heterozygous G>A transition (arrow) at position c.8815 of the cDNA sequence in individual II-6, family 2, as well as a heterozygous C>T transition (arrow) at position c.12250 of the cDNA sequence in individual II-2, family 4 (middle panels). The WT sequences (WT/WT) are given for comparison (lower panels). (B) Mildly affected individuals featured plantar involvement only while more severe cases carry variants in HMCN1 and showed at the same age, severe involvement of the hands and additional skin areas. (C) The predicted amino acid changes and the location of the three variants are depicted along a schematic representation of hemicentin-1 and its domains (SP, signal peptide; H, hemicentin domain; IGLR, tandem Ig-like repeat; TSP1, thrombospondin type1 repeat domain; G2F, G2 fragment domain; FC, fibulin carboxy-terminal domains; modified from Xu et al. [2013]).

Figure 2.

Protein modeling, dynamics simulation, and functional studies. (A) Evolutionary conservation analysis mapped onto the predicted protein structures. (B) The effect of HMCN1 variants on protein stability was calculated by FoldX (ΔΔG). All three variants are predicted to destabilize the protein substantially. (C) A scheme of an Ig tandem used for MD simulations is presented (left panel). Each of the Ig domains bearing a variant along with the two flanking Ig domains was modeled (with and without the deleterious variant) and used to run 250-ns MD simulations while measuring the interdomain angle as indicated in the left panel. The frequency distribution of the interdomain angle along the simulation trajectories is shown. Black curves denote the WT proteins, while the red curves denote the mutants as indicated. Note that the amino acid substitution causes significant deviations from the straight axis of the constructs harboring the Ig38 and Ig40 Ig domains compared to WT. (D) Structural studies of hemicentin-1 Ig27 domain are presented (each experiment was repeated three times). The left panel shows the elution profiles from size-exclusion chromatography for Ig27 WT (black) and mutant (red) domains. Inset, SDS-PAGE of purified Ig27-WT and Ig27-Gly2939Ser; molecular masses in kDa are indicated to the left. Note that the mutant domain shows an apparent larger molecular weight compared to WT. The right panel shows tryptophan fluorescence analysis of Ig27 WT (black) and mutant (red) domains, revealing a significant shift in emission wavelength and amplitude of the mutant domain as compared with the WT domain. (E) Purified Ig38 and Ig40 WT and mutants (p.Ala3969Thr and p.His4084Tyr, respectively) domains were analyzed by size exclusion chromatography (each experiment was repeated three times). Inset, SDS-PAGE analysis; molecular masses in kDa are indicated to the left. Note the reduced amount of Ig38 and absent Ig40 protein, as corroborated by the nickel exclusion chromatography data (Fig. S1). Source data are available for this figure: SourceData F2.

Figure S1.

Ig40 nickel exclusion chromatography. Elution profile from nickel exclusion chromatography experiment for WT and mutant Ig40 domain. A peak corresponding to purified protein appears only in WT protein (black arrow). The result represents three independent experiments.

Figure S2.

HMCN1 antibody specificity. (A) HaCaT cells were transfected with HMCN1-specific (HaCaT siHMCN1) or scramble (HaCat siControl) siRNAs. HMCN1 mRNA levels were quantified using RT-qPCR. Results were normalized to GAPDH mRNA, represent the mean ± SE of two experiments, and are expressed as a percentage of HMCN1 mRNA expression in cells transfected with siControl (two-sided t test; ***P < 0.005). (B) HaCaT cells were transfected with HMCN1-specific (siHMCN1) or scramble (siControl) siRNAs as described in A and immunostained using rabbit anti-hemicentin-1 antibody as described in Materials and methods. The negative control consisted of cells stained with the secondary antibody only. Note the decrease in hemicentin-1 staining in cells transfected with siHMCN1 (scale bar = 10 μm, hemicentin-1 [HMCN1] – red). (C) Fluorescence staining intensity in the experiment depicted in B was quantified by ImageJ. Result represents mean ± SE of two independent experiments (two-sided t test; *P < 0.05).

Figure 3.

Hemicentin-1 expression in the skin. (A) Skin biopsies obtained from a healthy individual were co-immunostained using anti-hemicentin-1 and anti-desmoglein-1 antibodies. Scale bars = 100 μm (left panel) and 20 μm (right panel); hemicentin-1 (HMCN1), green; desmoglein-1 (DSG1), red; nuclei are stained in blue by DAPI; the experiment was repeated three times. (B) Super-resolution fluorescence microscopy (dSTORM) pictures of skin biopsies obtained from a healthy individual were co-immunostained using anti-hemicentin-1 and anti-desmoglein 1 antibodies (scale bars = 1 μm [middle panel] and 0.6 μm [right panel]; hemicentin-1 [HMCN1], green; desmoglein-1 [DSG1], red). See corresponding videos (Videos 1 and 2) (the experiment was repeated two times).

Figure S3.

Immunofluorescence studies in human skin and in three-dimensional skin equivalents. (A) Skin sections obtained from a healthy individual were co-immunostained using anti- hemicentin-1 and anti-C17 antibodies (scale bars = 25 μm [left panel], 10 μm [middle panel], and 5 μm [right panel]; hemicentin-1 [HMCN1], green; C17, red; nuclei are stained in blue by DAPI). The experiment was repeated twice. (B–D) Human primary keratinocytes and fibroblasts transfected with HMCN1-specific siRNA (siHMCN1) or control siRNA (siControl) were used to generate three-dimensional organotypic cell cultures. (B) Punch biopsies were obtained from the skin equivalents at day 12, and co-stained for K14 and hemicentin-1 (HMCN1) (scale bar = 25 μm). (C) K14 and hemicentin-1 expression levels were quantified by ImageJ. Result represents mean ± SE of two independent experiments (one-way ANOVA; **P < 0.01, ***P < 0.005). (D) Human primary keratinocytes and fibroblasts transfected with HMCN1-specific siRNA (siHMCN1) or control siRNA (siControl) were used to generate three-dimensional organotypic cell cultures. Punch biopsies were obtained from the skin equivalents at day 12 and stained for C17 (scale bar = 25 μm). The experiment was repeated twice.

Figure 4.

Hemicentin-1 directly interacts with K 14. (A) A solid growth assay was used to validate hemicentin-1–K14 interaction identified in the Y2H assay. Yeasts were transfected with pB29-HMCN1 (aa 1–212, hemicentin-1) bait vector and pP7-KRT14 (aa 92–188, K14) prey vector. Yeast colonies were grown on DO-3 media (without tryptophan, leucine, and histidine), which selects for yeast harboring interacting bait and prey proteins. Note the presence of colonies in yeasts transfected with both constructs only. (B) Flag-tagged K14 and His-tagged hemicentin-1 VWFA domain-expressing plasmids were transfected into HeLa cells. Hemicentin-1 VWFA domain and K14 were co-immunoprecipitated using anti-Flag antibody-conjugated magnetic beads followed by immunoblotting of the precipitated proteins using anti-His antibody. The experiment was repeated three times with similar results. (C) PLA was performed using antibodies directed against hemicentin-1 and K14 in healthy human skin sections. Red dots indicate positive interaction (the experiment was repeated two times). The experiment was repeated using a skin biopsy taken from individual II-2, family 4 (scale bar, 50 μm), or kidney tissue a negative control. In addition, a PLA assay using antibodies directed against keratin 5 (K5) and K14 served as a positive control (scale bars, 25 μm). Source data are available for this figure: SourceData F4.

Figure S4.

Hemicentin-1 VWA domain, K14-encoding cDNA constructs expression in HeLa cells and HMCN1 silencing efficiency. (A) Hemicentin-1 VWA domain and K14 (KRT14)-expressing cDNA constructs were transfected in HeLa cells. Gene expression was quantified using RT-qPCR. Results were normalized to GAPDH mRNA, represent the mean ± SE of two experiments and are expressed as fold change of mRNA expression in cells transfected with an empty vector (one-way ANOVA; **P < 0.01, ***P < 0.005). (B)HMCN1 mRNA levels were quantified by RT-qPCR in primary human keratinocytes (KC) or fibroblasts (FB) transfected with HMCN1-specific (siHMCN1) or scramble (siControl) siRNAs. Results were normalized to GAPDH mRNA, represent the mean ± SE of three experiments, and are expressed as a percentage of HMCN1 mRNA expression in cells transfected with siControl (two-way t test; ***P < 0.005).

Figure 5.

Hemicentin-1 deficiency effect on KIF organization. (A) HaCaT cells grown to 60% confluence on glass coverslips in 12 wells plates were downregulated for HMCN1 with a specific siRNA (siHMCN1) or with a control siRNA (siControl) and co-transfected with an empty vector (EV) or expression vectors encoding either, WT cDNA, or mutant (Mut) KRT14 cDNA carrying the c.373C>T, p.Arg125Cys variant for 48 h at 37°C. The cells were stained for KRT14 expression (red staining) and DAPI (blue staining) (the experiment was repeated two times). Cells were visualized by confocal microscopy. Note protein aggregates in cells transfected with the mutant KRT14 as well as lower expression levels of K14 in siHMCN1-transfected cells (scale bar = 10 μm). (B) K14 KIF fluorescence intensity per cell was measured using ImageJ, analyzing four different zones for each condition. Statistical analysis was performed by t test (***P < 0.005; ****P < 0.001).

Figure S5.

Hemicentin-1 deficiency effect on KIF organization in primary keratinocytes. Primary keratinocytes cells grown to 60% confluence on glass coverslips in 12 wells plates were downregulated for hemicentin 1 with a specific siRNA (siHMCN1) or with a control siRNA (siControl), and co-transfected with an empty vector (EV) or expression vectors encoding either WT cDNA or mutant (Mut) KRT14 cDNA carrying the c.373C>T, p.Arg125Cys variant for 48 h at 37°C. The cells were stained for KRT14 expression (red staining) and DAPI (blue staining). The experiment was repeated twice. Cells were visualized by confocal microscopy. Note decreased K14 expression upon HMCN1 silencing and protein aggregates in cells transfected with the mutant KRT14, with marked aggravation in siHMCN1-transfected cells (scale bar = 10 μm).

Figure 6.

Three-dimensional modeling of hemicentin-1 deficiency. (A) Human primary keratinocytes and fibroblasts transfected with HMCN1-specific siRNA (siHMCN1) or control siRNA (siControl) were used to generate three-dimensional organotypic cell cultures. Punch biopsies were obtained from the skin equivalents at day 12 and stained for hematoxylin and eosin (scale bar, 50 μm). Red arrows indicate subepidermal blisters. (B) Subepidermal blistering was quantitatively ascertained as the ratio of the length of the blistering BMZ divided by the entire length of the BMZ using the NIS-Elements BR 3.2 software (Nikon). The plot represents the results of two independent experiments in which a total of 58 random fields were examined (two-sided t test; ***P < 0.005).