Desmoplakin Variants Are Associated with Idiopathic Pulmonary Fibrosis

Abstract

Rationale: Sequence variation, methylation differences, and transcriptional changes in desmoplakin (DSP) have been observed in patients with idiopathic pulmonary fibrosis (IPF).

Objectives: To identify novel variants in DSP associated with IPF and to characterize the relationship of these IPF sequence variants with DSP gene expression in human lung.

Methods: A chromosome 6 locus (7,370,061-7,606,946) was sequenced in 230 subjects with IPF and 228 control subjects. Validation genotyping of disease-associated variants was conducted in 936 subjects with IPF and 936 control subjects. DSP gene expression was measured in lung tissue from 334 subjects with IPF and 201 control subjects.

Measurements and main results: We identified 23 sequence variants in the chromosome 6 locus associated with IPF. Genotyping of selected variants in our validation cohort revealed that noncoding intron 1 variant rs2744371 (odds ratio = 0.77, 95% confidence interval [CI] = 0.66-0.91, P = 0.002) is protective for IPF, and a previously described IPF-associated intron 5 variant (rs2076295) is associated with increased risk of IPF (odds ratio = 1.36, 95% CI = 1.19-1.56, P < 0.001) after controlling for sex and age. DSP expression is 2.3-fold increased (95% CI = 1.91-2.71) in IPF lung tissue (P < 0.0001). Only the minor allele at rs2076295 is associated with decreased DSP expression (P = 0.001). Staining of fibrotic and normal human lung tissue localized DSP to airway epithelia.

Conclusions: Sequence variants in DSP are associated with IPF, and rs2076295 genotype is associated with differential expression of DSP in the lung. DSP expression is increased in IPF lung and concentrated in the airway epithelia, suggesting a potential role for DSP in the pathogenesis of IPF.

Keywords: desmoplakin; genetics; idiopathic interstitial pneumonia; idiopathic pulmonary fibrosis; intron.

Figure 1.

Desmoplakin (DSP) variants and association with idiopathic pulmonary fibrosis (IPF). This figure illustrates the relative position in the DSP gene of the variants genotyped in this study. rs2076295 is the intron 5 variant previously described by Fingerlin and colleagues (4) to be associated with fibrosing idiopathic interstitial pneumonias, of which IPF is the most common. The variants in red were significantly associated with IPF when controlling for sex and age (P < 0.01). Odds ratios (ORs) are presented with 95% confidence intervals noted in parentheses.

Figure 2.

Desmoplakin (DSP) variants and association with idiopathic pulmonary fibrosis (IPF) when controlling for rs2076295. This figure illustrates the relative position in the DSP gene of the variants genotyped in this study. rs2076295 is the intron 5 variant previously described by Fingerlin and colleagues (4) to be associated with fibrosing idiopathic interstitial pneumonias, of which IPF is the most common. The variant in red (rs2744371) is nominally associated with IPF when controlling for age, sex, and rs2076295 genotype (P < 0.05). Odds ratios (ORs) are presented with 95% confidence intervals noted in parentheses.

Figure 3.

Relative desmoplakin (DSP) expression in human lung tissue by quantitative polymerase chain reaction, grouped by diagnosis. Idiopathic pulmonary fibrosis (IPF) lung tissue had significantly higher DSP expression than control lung. Statistical tests (two-tailed Student’s t test) were performed on ∆ cycle threshold (C_t) values (DSP C_t values normalized to housekeeping gene GAPDH), and relative expression is visually depicted as 10 − ∆C_t. Means are indicated within each group with a horizontal line; error bars indicate SEM. UIP = usual interstitial pneumonia.

Figure 4.

Desmoplakin (DSP) expression and age. In control lung samples, DSP expression falls with age (P = 0.004) (A), but in idiopathic pulmonary fibrosis lungs, DSP expression remains constant with increasing age (B). Data points are plotted as green circles; the line of best fit is also plotted with 95% confidence interval indicated by gray areas. Relative DSP expression is plotted on the y-axis (10 − ∆ cycle threshold [C_t]), and age is plotted on the x-axis. Statistical analysis was performed using linear regression on ∆C_t values. CI = confidence interval.

Figure 5.

Relative desmoplakin (DSP) expression by rs2076295 genotype. DSP expression as measured by quantitative polymerase chain reaction (normalized to GAPDH) varied by rs2076295 genotype. When lung samples were divided into idiopathic pulmonary fibrosis (A) and control (B) groups, there were significant differences between DSP gene expression by genotype at rs2076295. Means are indicated within each group with a horizontal line; error bars indicate SEM. Statistical tests were performed on raw ∆ cycle threshold (C_t) values for each group; relative DSP expression is depicted visually as 10 − ∆C_t. ANOVA = analysis of variance.

Figure 6.

Immunohistochemistry of desmoplakin (DSP) in normal and fibrotic idiopathic pulmonary fibrosis (IPF) lungs. Immunohistochemical staining of formalin-fixed, paraffin-embedded human lung tissue with rabbit polyclonal antibody against DSP I/II (Santa Cruz Biotechnology, Inc., Dallas, TX). Normal lung stained for DSP shows strong staining along the airway epithelia (A), in the trachea (C), and in the more distal airways (D) without significant alveolar staining (A). IPF lung stained concurrently shows strong staining of airway epithelia in both normal-appearing and fibrotic regions (B and E). Cystic areas of the fibrotic lung are lined with epithelial cells that strongly stain positive for DSP I/II (F).