Adam19 Deficiency Impacts Pulmonary Function: Human GWAS Follow-up in Mouse

Abstract

Purpose: Over 550 loci have been associated with human pulmonary function in genome-wide association studies (GWAS); however, the causal role of most remains uncertain. Single nucleotide polymorphisms in a disintegrin and metalloprotease domain 19 (ADAM19) are consistently related to pulmonary function in GWAS. Thus, we used a mouse model to investigate the causal link between Adam19 and pulmonary function.

Methods: We created an Adam19 knockout (KO) mouse model and validated the gene targeting using RNA-Seq and RT-qPCR. Contrary to prior publications, the KO was not neonatal lethal. Thus, we phenotyped the Adam19 KO.

Results: KO mice had lower body weight and shorter tibial length than wild type (WT). Dual-energy X-ray Absorptiometry indicated lower soft weight, fat weight, and bone mineral content in KO mice. In lung function analyses using flexiVent, compared to WT, Adam19 KO had decreased baseline respiratory system elastance, minute work of breathing, tissue damping, tissue elastance, and forced expiratory flow at 50% forced vital capacity but higher FEV_0.1 and FVC. Adam19 KO had attenuated tissue damping and tissue elastance in response to methacholine following LPS exposure. Adam19 KO also exhibited attenuated neutrophil extravasation into the airway after LPS administration compared to WT. RNA-Seq analysis of KO and WT lungs identified several differentially expressed genes (Cd300lg, Kpna2, and Pttg1) implicated in lung biology and pathogenesis. Gene set enrichment analysis identified negative enrichment for TNF pathways.

Conclusion: Our murine findings support a causal role of ADAM19, implicated in human GWAS, in regulating pulmonary function.

Keywords: RNA-Seq; flexiVent; inflammation; lung function; meltrin beta; spirometry.

Figure 1. Gene targeting scheme (A) for Adam19-tdTomato allele and validation by RNA-Seq (B).

A. Adam19 WT: Endogenous wild-type locus. Adam19 Flox: Adam19 conditional null (“flox”) allele with exon 6 and 7 floxed by LoxP sites (solid yellow triangles). Adam19-tdTomato: Adam19 mutant allele in which exons 6 and 7 are replaced by the tdTomato construct, disrupting Adam19 gene expression. Homozygosity for Adam19-tdTomato alleles is equivalent to Adam19 knockout (KO). Each blue box represents an exon; the exon number is underneath. pA: polyA; hygroR: Hygromycin Resistance. CAG: CMV enhancer, chicken beta-Actin promoter, and rabbit beta-Globin splice acceptor site. B. Read densities of Adam19 exons and junctions in WT and KO mice by RNA-Seq analysis. The blue boxes represent exons. In the Adam19 KO, exons 6 and 7 were replaced by the tdTomato construct and showed minimal transcript expression from exon 8 through the end of the Adam19 gene. Bold orange or bold purple regions represent aggregate sequence read depth across the Adam19 WT (left panel) or KO (right panel) gene locus. Light orange or light purple arcs indicate sequence reads whose alignments represent observed splice junctions. The arc width indicates the aggregate number of junction reads. No in-frame splicing events were detected from exon 5 to any downstream Adam19 exons. n=3 mice per genotype per heart or lung tissue.

Figure 2. Adam19 KO mice have reduced body weight, shorter tibia length, and altered body composition.

A. Body weight measured using a top-loading scale (WT: n=114, KO: n=104). B. Tibia length measured using a ruler (WT: n=24, KO: n=20). C: Body weight of mice for measuring body composition. D-L: Body composition parameters obtained using Dual-energy X-ray Absorptiometry (WT: n=10, KO: n=9). Total weight = soft weight + bone mineral content (BMC), soft weight = lean weight + fat weight, Fat % = fat weight/soft weight in percentage, BMD = bone mineral density = BMC/bone area. p values < 0.05 for differences in the parameter by genotype are displayed (A-L).

Figure 3. Adam19 deficiency alters (A) baseline mechanics and (B) spirometry parameters determined by flexiVent.

n=22 for WT, n=15 for KO. R_rs=resistance of the respiratory system; E_rs=elastance of the respiratory system; mWOB=minute work of breathing (the work required to breath-in on a minute basis); J=joule (one joule is the work required to move 1 liter of gas through a 10-cmH₂O pressure gradient). R_N=Newtonian resistance; G=tissue damping; H=tissue elastance; FEV_0.1=forced expiratory volume in 0.1 s; FVC=forced vital capacity; FEV_0.1/FVC = the ratio of FEV_0.1 over FVC in %; FEF50=Forced expiratory flow at 50% FVC; PEF=Peak expiratory flow; FEV_PEF=Forced expiratory volume at peak expiratory flow. p<0.05 for differences in the parameter by genotype are displayed.

Figure 4. Adam19 deficient mice have reduced airway responsiveness to methacholine following LPS exposure.

The maximum response to methacholine at each dose was expressed as a percentage of the maximum response at PBS. Means and standard errors of means are indicated as bar lines. R_rs=resistance of the respiratory system; E_rs=elastance of the respiratory system; mWOB=minute work of breathing; R_N=Newtonian resistance; G=tissue damping; H=tissue elastance; BL=baseline; PBS=phosphate buffered saline. *p values ≤ 0.05 are shown for the genotype difference of the slope difference of the response to methacholine following LPS exposure (vs. saline). n=13 for WT-Saline, n=20 for WT-LPS, n=14 for KO-Saline, n=21 for KO-LPS

Figure 5. Analysis of differential immune cell counts in bronchoalveolar lavage fluid (BALF) in mice following LPS exposure.

Increases in neutrophil number (C) induced by LPS (vs. saline) were lower in Adam19 KO than in WT mice (46% fewer cells, p=0.032). The degree of increase in total cells (A) following LPS (vs. saline) was lower in KO (the ratio of the LPS effect between the KO and WT = 0.61, 95% confidence interval (CI) = 0.36–1.02, p=0.058). No genotype differences of LPS effects were identified for macrophages (B) and lymphocytes (D). SAL=saline, WT: n=10 (SAL), 10 (LPS); KO: n=7 (SAL), 8 (LPS).