Integrated study of systemic and local airway transcriptomes in asthma reveals causal mediation of systemic effects by airway key drivers

Abstract

Background: Systemic and local profiles have each been associated with asthma, but parsing causal relationships between system-wide and airway-specific processes can be challenging. We sought to investigate systemic and airway processes in asthma and their causal relationships.

Methods: Three hundred forty-one participants with persistent asthma and non-asthmatic controls were recruited and underwent peripheral blood mononuclear cell (PBMC) collection and nasal brushing. Transcriptome-wide RNA sequencing of the PBMC and nasal samples and a series of analyses were then performed using a discovery and independent test set approach at each step to ensure rigor. Analytic steps included differential expression analyses, coexpression and probabilistic causal (Bayesian) network constructions, key driver analyses, and causal mediation models.

Results: Among the 341 participants, the median age was 13 years (IQR = 10-16), 164 (48%) were female, and 200 (58.7%) had persistent asthma with mean Asthma Control Test (ACT) score 16.6 (SD = 4.2). PBMC genes associated with asthma were enriched in co-expression modules for NK cell-mediated cytotoxicity (fold enrichment = 4.5, FDR = 6.47 × 10^-32) and interleukin production (fold enrichment = 2.0, FDR = 1.01 × 10^-15). Probabilistic causal network and key driver analyses identified NK cell granule protein (NKG7, fold change = 22.7, FDR = 1.02 × 10^-31) and perforin (PRF1, fold change = 14.9, FDR = 1.31 × 10^-22) as key drivers predicted to causally regulate PBMC asthma modules. Nasal genes associated with asthma were enriched in the tricarboxylic acid (TCA) cycle module (fold enrichment = 7.5 FDR = 5.09 × 10^-107), with network analyses identifying G3BP stress granule assembly factor 1 (G3BP1, fold change = 9.1 FDR = 2.77 × 10^-5) and InaD-like protein (INADL, fold change = 5.3 FDR = 2.98 × 10^-9) as nasal key drivers. Causal mediation analyses revealed that associations between PBMC key drivers and asthma are causally mediated by nasal key drivers (FDR = 0.0076 to 0.015).

Conclusions: Integrated study of the systemic and airway transcriptomes in a well-phenotyped asthma cohort identified causal key drivers of asthma among PBMC and nasal transcripts. Associations between PBMC key drivers and asthma are causally mediated by nasal key drivers.

Keywords: Airway; Asthma; Blood; Causal network; Interleukin; NK cell; Nasal; Peripheral blood mononuclear cell; Transcriptome; Tricarboxylic acid.

Fig. 1

Study flow. Peripheral blood mononuclear cell (PBMC) and nasal transcriptome profiles from 341 participants with and without asthma were generated and studied to characterize systemic and airway responses in asthma. Using PBMC transcriptome data from the discovery set, we first identified differentially expressed genes (DEGs) associated with asthma (yellow box). Genes that were also associated with asthma with the same direct of effect in the independent test set were deemed validated “PBMC asthma genes.” Weighted gene coexpression network analysis and enrichment testing were then performed to identify co-expression modules enriched for PBMC asthma genes (PBMC asthma modules). Next, probabilistic causal (Bayesian) networks were built separately for the discovery set and test set. Key driver analysis was performed on each network using PBMC asthma module members as targets. Key drivers identified in both the discovery and test sets were deemed “PBMC key drivers.” The same series of analyses was performed with the nasal transcriptome data generated in parallel from participants to identify nasal asthma genes, nasal asthma modules, and nasal key drivers (green box). To characterize relationships between the PBMC key drivers and nasal key drivers identified, causal mediation analyses were then performed (purple box)

Fig. 2

PBMC genes associated with asthma. Differential gene expression analysis was performed. Genes associated with asthma in the discovery set (FDR $\le$ 0.05) are shown in orange, and those that were also associated with asthma with the same direction of effect in the independent test set were deemed validated PBMC asthma genes. Key drivers identified and validated in downstream analyses are labeled, including key drivers of the NK cell-mediated cytotoxicity module (purple) and key drivers of the interleukin production module (blue)

Fig. 3

Probabilistic causal network and functional biological context for the NK cell-mediated cytotoxicity PBMC module. A Probabilistic causal network and key driver analysis results for the NK cell-mediated cytotoxicity module. This PBMC transcriptome module was significantly enriched with PBMC asthma genes. The arrow indicates the overall causality flow with key drivers at the top level. Level indicates path length of the gene from a key driver. Genes on higher levels have greater causal impact on downstream genes. Color, shade, and shape indicate key driver, module membership, and PBMC asthma genes as summarized in the legend. Selected non-key drivers are additionally labeled, as they are recognized to function with the upstream key drivers. B Functional biological context for the NK cell-mediated cytotoxicity module. Genes within purple boxes are key drivers and genes in purple font are module genes highlighted in A. Dashed arrows indicate causal relationships inferred from the causal network. NKG7 encodes natural killer cell granule protein 7, which regulates granule exocytosis in lymphocytes [48]. PRF1 encodes perforins that function with granzymes and granulysins to kill target cells [47]. Killer cell lectin-like receipt D1 encoded by KLRD1 forms heterodimers with NKG2 and can stimulate or inhibit cytotoxicity depending on NKG2 isoform [58, 59]. MBYL1, MYB proto-oncogene like 1, is a transcription activator [53]. The module genes GZMB, GZMA and GNLY encode granzymes and granulysin that kill target cells upon release [47, 49]

Fig. 4

Probabilistic causal network and functional biological context for the interleukin production PBMC module. A Probabilistic causal network and key driver analysis results for the interleukin production module. This PBMC transcriptome module was significantly enriched with PBMC asthma genes. The arrow indicates the overall causality flow with the key drivers on the top level. Level indicates path length of the gene from a key driver. Genes on higher levels have greater causal impact on downstream genes. Color, shade, and shape indicate key driver, module membership, and PBMC asthma genes as summarized in the legend. Some genes downstream of key drivers are additionally highlighted given their recognized roles in immune-related functions. B Functional biological context for the interleukin production module. Genes highlighted within blue boxes are key drivers of the module. (1) CTSS, Cathepsin S, is a lysosomal cysteine proteinase that cleaves off the invariant chain on MHC class II molecules in the endolysosomal compartments for later antigen-MHC II formation [60]. ATP6AP2 encodes a protein involved in lysosomal proton-transporting V-type ATPase [55]. (2) TNFSF13 encodes a member of the tumor necrosis factor ligand superfamily important for B cell development [61]. (3) RAB3D encodes a member of the RAS oncogene family that regulates secretory granule maturation [56]. (4) PSAP encodes prosaposin, which yields Saposin B when cleaved [50]. Saposin B works with other enzymes to break down sphingolipids [52]. (5) CAPZA2, capping actin protein of muscle Z-line subunit alpha 2, is an F-actin capping protein that caps the barbed end of actin filaments [57]. (6) LCP1 encodes L-plastins (LPL) that bind F-actin. LPL also mediates sensitization of eosinophils [51]

Fig. 5

Associations between PBMC key drivers and asthma are causally mediated by nasal key drivers. PBMC key drivers and nasal key drivers associated with asthma were examined. Triangle edges summarize pairwise associations. Red arrows indicate significant causal mediation by nasal key drivers (FDR $\le$ 0.05)