Host-pathogen interaction profiling of nontypeable Haemophilus influenzae and Moraxella catarrhalis coinfection of bronchial epithelial cells

Abstract

Chronic obstructive pulmonary disease (COPD) is a progressive inflammatory lung disease and the third leading cause of death globally. Nontypeable Haemophilus influenzae (NTHi) and Moraxella catarrhalis (Mcat) are common pathogens in individuals with COPD. Acquisition of NTHi or Mcat can cause acute exacerbations of COPD. NTHi and/or Mcat also persist for months in the lower airways and lead to chronic inflammation. We hypothesized that infection by NTHi or Mcat, singly or during coinfection, requires regulation of specific bacterial and host cell pathways. We investigated this phenomenon using an in vitro cell culture model consisting of lung carcinoma H292 cell lines, infected with NTHi, Mcat, or both species. Samples were fractionated into "apical fluid," containing free-floating bacteria, and adhered/invaded bacteria on or within H292 cells. We used transcriptomic profiling with RNA-seq and various bioinformatic analyses to disentangle host-pathogen interactions in epithelial cell infection from the perspective of each species. Several biological pathways were differentially regulated across all conditions (31, NTHi; 22, Mcat; and 169, human). NTHi transcriptomic profiles differed during mono-infection and coinfection; examples included downregulation of iron-sulfur metabolism (IscR regulon) and differential regulation of quorum sensing in coinfection compared to mono-infection. Mcat was comparatively less affected by the presence of NTHi during coinfection. H292 epithelial cells responded broadly to all infections with distinct responses to mono-infection and coinfection. Enriched host pathways included influenza/interferon/Wnt and proinflammatory responses. These findings suggest common and distinct processes involved in NTHi and/or Mcat-induced COPD pathogenesis and have implications for therapeutic intervention.IMPORTANCEChronic obstructive pulmonary disease (COPD) is a leading cause of death worldwide. Bacteria such as nontypeable Haemophilus influenzae (NTHi) and Moraxella catarrhalis (Mcat) can cause exacerbations of COPD, and they can persist in the lungs for months, which increases inflammation. We studied how these bacteria interact with lung cells by infecting a cell culture model with NTHi, Mcat, or both. We used RNA sequencing and bioinformatic analysis to examine how the bacteria and host cells respond. When NTHi and Mcat were present together, they behaved differently than when each was alone. We found that different host biological pathways were activated during infection, including those related to inflammation and immune responses. These results provide insights into how NTHi and Mcat contribute to COPD progression and suggest potential targets for new treatments.

Keywords: Haemophilus influenzae; Moraxella catarrhalis; RNA sequencing; chronic obstructive pulmonary disease; coinfection.

Fig 1

Overview of NTHi and Mcat infection or coinfection of human bronchial epithelial cells (H292) for multi-species RNA-seq. NTHi and Mcat were grown individually or co-cultured in H292 cell culture media in the absence of H292 cells (top three labeled flasks) as bacterial control samples. H292 cells were grown in the absence of bacteria (uninfected control, far left flask), infected individually with NTHi or Mcat, or coinfected (Mix) with both bacteria (bottom four flasks). H292 infected samples had apical fluid (“apf” samples), containing free-floating non-adhered/non-invaded bacteria, and infected H292 cells, with adhered/invaded bacteria on or within H292 cells (“cell” samples), harvested and sequenced separately. Additionally, bacteria-free Mcat apical fluid (Msup) was introduced into NTHi-infected H292 cells to determine Mcat secretory effects (far right flask), referred to as Mcat secretome. All samples were cultured for 5 hours, except NTHi-infected samples with Mcat apical fluid, which had 5 hours of additional growth. Created in BioRender.

Fig 2

PCA of NTHi, Mcat, and H292 samples. (A) PCA of NTHi samples. PC1 (~45% of the variance) separates the infection conditions with NTHi adhered/invaded to H292 cells (cell) on the left and apf or individual NTHi samples (NTHi) on the right. PC2 (19%) partially separates coinfection (Mix) samples (bottom, contain Mcat) from just NTHi samples (top). (B) PCA of Mcat samples. PC1 (63%), similar to NTHi, separates Mcat adhered/invaded to H292 cells (left) from apf or individual Mcat samples (right). PC2 (11%) shows no clear trend. (C) PCA of H292 samples. PC1 (~44%) separates all infected cells (left) from uninfected cells (right). PC2 (~15%) separates each different infection from top to bottom. Normalized expression values (VST counts) used in each PCA are provided in Tables S2 to S4.

Fig 3

Differential expression analysis of NTHi, Mcat, and human genes. (A) Counts (number) of NTHi DE genes across various comparisons. Eighty-six genes were DE across samples with H292 cells vs those without. (B) Counts of Mcat DE genes across various comparisons. Eighty-three genes were found to be DE across samples with H292 cells vs those without. (C) Counts of human DE genes across infected vs uninfected samples. A total of 3,240 genes were shared across all four comparisons. Lists of all DE genes for each comparison for each species are provided in Table S2 (NTHi), Table S3 (Mcat), and Table S4 (human).

Fig 4

Condition-specific gene expression patterns of NTHi, Mcat, and human DE genes. (A) WGCNA heatmap of normalized expression levels (VST counts) of NTHi DE genes showing eight modules (NTHi_A–NTHi_H) which correlated with specific conditions. Black boxes show NTHi gene expression patterns associated with epithelial cell infection. (B) WGCNA heatmap of normalized expression levels of Mcat DE genes showing six modules (Mcat_A–Mcat_F) which correlated with specific conditions. (C) WGCNA heatmap of normalized expression levels of human DE genes showing two modules which correlated with Mcat infected (H292_A) and NTHi infected (H292_B) samples. A list of DE genes in each module is provided in Table S5 (NTHi and Mcat) and Table S6 (human).

Fig 5

Pathway enrichment analysis based on NTHi, Mcat, and human DE genes. X axes represent the percentage of detected genes in the pathway, and Y axes represent the enriched pathways and the number of genes that were detected. Percentage indicates the fraction of DE genes detected relative to all genes of a pathway. (A) Enrichment statistics of all significantly enriched pathways of NTHi DE genes (KEGG and RegPrecise databases). (B) Enrichment statistics of all significantly enriched pathways of Mcat (KEGG and RegPrecise). (C) Enrichment statistics of all significantly enriched pathways of human DE genes (KEGG via ShinyGO). Expanded lists of enriched pathways are provided in Table S5 (NTHi and Mcat) and Table S6 (human).

Fig 6

Proposed model of NTHi and Mcat coinfection. Green arrows and text represent NTHi-induced changes or host responses to NTHi. Purple arrows and text represent Mcat-induced changes or host responses to Mcat. Up or down arrow glyphs indicate gene up or downregulation identified in this study. Bullet points indicate enriched gene sets or pathways. Abbreviation: NLRs, nucleotide-binding oligomerization domain-like receptors. Created in BioRender.