Modular Transcriptional Networks of the Host Pulmonary Response during Early and Late Pneumococcal Pneumonia

Abstract

Streptococcus pneumoniae (Spneu) remains the most lethal bacterial pathogen and the dominant agent of community-acquired pneumonia. Treatment has perennially focused on the use of antibiotics, albeit scrutinized due to the occurrence of antibiotic-resistant Spneu strains. Immunomodulatory strategies have emerged as potential treatment options. Although promising, immunomodulation can lead to improper tissue functions either at steady state or upon infectious challenge. This argues for the availability of tools to enable a detailed assessment of whole pulmonary functions during the course of infection, not only those functions biased to the defense response. Thus, through the use of an unbiased tissue microarray and bioinformatics approach, we aimed to construct a comprehensive map of whole-lung transcriptional activity and cellular pathways during the course of pneumococcal pneumonia. We performed genome-wide transcriptional analysis of whole lungs before and 6 and 48 h after Spneu infection in mice. The 4,000 most variable transcripts across all samples were used to assemble a gene coexpression network comprising 13 intercorrelating modules (clusters of genes). Fifty-four percent of this whole-lung transcriptional network was altered 6 and 48 h after Spneu infection. Canonical signaling pathway analysis uncovered known pathways imparting protection, including IL17A/IL17F signaling and previously undetected mechanisms that included lipid metabolism. Through in silico prediction of cell types, pathways were observed to enrich for distinct cell types such as a novel stromal cell lipid metabolism pathway. These cellular mechanisms were furthermore anchored at functional hub genes of cellular fate, differentiation, growth and transcription. Collectively, we provide a benchmark unsupervised map of whole-lung transcriptional relationships and cellular activity during early and late pneumococcal pneumonia.

Figure 1

Immunohistochemical analysis of the early and late host pulmonary response to pneumococcal infection. C57BL/6 female mice were each intranasally inoculated with 2 × 10⁷ CFUs S. pneumoniae D39. Mice were killed at 6 or 48 h (n = 4 each) and lungs were used for histological examination. (A) HE-stained representative lung sections of noninfected (C57BL/6; n = 3), 6-h postinfection (Spneu 6 h) and 48-h postinfection (Spneu 48 h) mice. (B) Histological scoring, determined as described in Materials and Methods, showed inflammatory pathology was largely similar between 6 h and 48 h. (C, D) Granulocyte abundance was higher at 6 h postinfection than at 48 h postinfection, evidenced by higher anti-Ly6G+ cell counts. (E) Bacterial burden measured as CFUs/mL was higher at 6 h postinfection than at 48 h postinfection. Data represent means ± SEM.

Figure 2

Genome-wide pulmonary gene expression analysis during early and late pneumococcal pneumonia. (A, B) Volcano plot representation (integrating probabilities and fold changes) of the transcriptional changes that occur at (A) 6 h postinfection (n = 4) and (B) at 48 h postinfection (n = 4), compared to noninfected controls (n = 3). Horizontal red line denotes BH-adjusted p < 0.05. (C) Heat-map representation of the 84 top differential genes (BH p < 0.01 and log₂ fold change ≥2 or ≤−2) determined by direct comparison of the 6- and 48-h postinfection samples. (D) Principal component analysis by decomposing the 84 top differential genes to 2 principal components resulted in a cumulative explainable variance of 95% (Dimension 1, 67%; Dimension 2, 28%).

Figure 3

Modular organization of the pulmonary transcriptome during pneumococcal pneumonia. The top 4,000 most variable genes (coefficient of variation) across noninfected 6-h and 48-h postinfection samples were used for construction of a weighted pairwise Spearman correlation matrix and clustered into 13 transcriptional modules, each encompassing more than 100 genes. (A) Weighted correlation coefficients were imported into the Cytoscape software (www.cytoscape.org) for unsupervised visualization of the network geometry given module memberships (color-coded) by yFiles layout. Genes with high correlation coefficients (weight >0.1) were used. Each module was labeled by ingenuity pathway analysis (IPA)-derived biological pathways (BH-adjusted Fisher p value <0.05). Based on this layout IL17A/IL17F signaling module (mod2) was central to the host pulmonary response. (B) Modules were significantly associated (BH p < 0.05) to shared and distinct cell-types that allowed for their stratification into myeloid, lymphoid or stromal cell enriched modules. Axis in radar graphs denotes the negative log₁₀ BH p-value. Note: 7 transcriptional modules yielded no significant association to known biological pathways, which may constitute as yet unexplored biological components of the host pulmonary response; these modules are indicated by their module number.

Figure 4

Kinetic behavior of host pulmonary transcriptional modules during pneumococcal pneumonia. (A) Logistic regression was used to assess the significance of module activity given early (Spneu 6 h) and late (Spneu 48 h) pneumococcal pneumonia in C57BL/6 mice. Significance was demarcated by BH-adjusted p < 0.05. Seven modules were significantly associated with the early host pulmonary response phase, namely mod1, IL17a/IL17F signaling, NF-κB signaling, complement system, lipid metabolism, interferon signaling and mod9. Four modules were significantly associated with the late host pulmonary response phase, namely, IL17A/IL17F signaling, NF-κB signaling, complement system and interferon signaling. Pie charts depict the proportion of module genes with log₂ fold changes <−1.5 or >1.5 that differed significantly (BH p < 0.05) at 6 h or 48 h postinfection, compared to noninfected mice. Note: 7 transcriptional modules yielded no significant association with known biological pathways; these modules are indicated by their module number. (B) Differential gene expression analysis was performed using publicly available data (GSE49533) of BALB/c mice, uninfected (n = 5) and 6 h after S. pneumoniae serotype 2 strain D39 intransal infection (n = 5). (C) Spearman correlation analysis of the BALB/c and C57BL/6 mouse pulmonary response (considering log₂ fold changes) after 6 h pneumococcal infection showed a strong correlation. (D and E) Weighted gene coexpression and logistic regression analyses of the 6-h BALB/c response revealed 14 modules of closely correlating genes, of which four modules were significantly conserved between mouse strains. Pie charts depict the proportion of BALB/c module genes with log₂ fold changes <−1.5 or >1.5 that differed significantly (BH p < 0.05) at 6 h postinfection, compared to noninfected mice.

Figure 5

To gain insight into the regulatory features of the host pulmonary modular transcriptome, driver genes were identified on the basis of their intramodular connectivity (number of correlating genes within the same module). Driver genes were highlighted in their respective module-specific volcano plots (integrating BH p values and fold changes) at both 6 h (Spneu 6 h) and 48 h (Spneu 48 h) postinfection. (A) IL17A/IL17F signaling module (mod2); (B) interferon signaling module (mod7); (C) NF-κB signaling module (mod3); (D) undetermined pathway module (mod1); (E) complement system module (mod5); (F) lipid metabolism module (mod6); (G) undetermined pathway module (mod9). Red dots denote genes presenting BH p < 0.05 and log₂ fold change >1.5; blue dots denote genes presenting BH p < 0.05 and log₂ fold change <−1.5.