Stratified Functional Analysis of Bacterial Genes Associated With Pulmonary Exacerbations in Cystic Fibrosis

Cystic fibrosis (CF) is a genetic disorder characterized by impaired mucociliary clearance, leading to chronic airway infections and inflammation. Pulmonary exacerbations (PEx), marked by acute decreases in lung function, necessitate antibiotic intervention. This cyclical process of infection, inflammation, damage, treatment, and reinfection creates a detrimental feedback loop. The genetic complexity of CF, along with variable external factors, makes universal treatment challenging. While CFTR modulators and broad-spectrum antibiotics are standard treatments, understanding the dynamic interplay of the CF airway microbiome across various clinical states might offer novel therapeutic insights. This study aimed to conduct a stratified functional analysis of bacterial genes at three distinct time points during PEx management (PEx onset, end of antibiotic treatment, and follow-up) to determine the role of specific airway microbiome members in each clinical state. A secondary objective was to correlate changes in clinical states with the metabolic activity of specific airway microbiome community members. The hypothesis was that differences in gene expression of specific airway community members are detectable between clinical states and correlate with differences in metabolic activity.

Existing literature highlights the crucial role of the airway microbiome in CF lung disease. Studies have shown fluctuations in airway bacterial communities associated with clinical states and disease stages in CF, emphasizing the dynamic nature of the microbiome and its potential impact on disease progression. Research indicates that the metabolome could be used to predict or diagnose pulmonary exacerbations, suggesting the potential for metabolic markers as diagnostic and prognostic tools. Other research has investigated the resistance and resilience of the respiratory microbiota to PEx and subsequent antimicrobial intervention, while further studies have explored the specific changes in the CF airway microbiota at the time of PEx. The role of anaerobic bacteria in the CF airway, particularly Veillonella, remains an area of active investigation, with studies reporting both positive and negative correlations between their presence and lung function. The impact of antibiotic treatment on microbiome diversity is also an important area of research.

This single-center, prospective observational study enrolled CF patients treated with intravenous antibiotics for PEx between 2016 and 2020 at Children's National Hospital. Respiratory samples (sputum, oropharyngeal swabs, or bronchoalveolar lavage) were collected at three time points: PEx onset, end of antibiotic treatment, and follow-up. Standard procedures were followed for sample processing, including homogenization and centrifugation. Bacterial DNA was extracted using the QIAamp DNA Microbiome Kit, quantified using Qubit, and quality-assessed using Bioanalyzer. Libraries were constructed using a Nextera XT Library Prep Kit and sequenced on a NextSeq 500. Bioinformatic analysis involved quality control using FastQC and Flexbar, human sequence removal using KneadData, and taxonomic assignment using MetaPhlAn3. HUMAnN3 was employed for functional profiling (gene, pathway analysis), utilizing MetaCyc for pathway collection. Statistical analysis in R included DESeq2 for differential abundance analysis, ggplot2 for visualization, phyloseq for microbiome data analysis, and vegan for community ecology analysis. PERMANOVA was used to assess beta diversity. Paired t-tests were used for comparisons of alpha diversity measures and ppFEV1 between time points. STATA/IC was used for additional statistical analyses.

Twenty-two CF patients experienced 45 PEx during the study period. 221 bacterial species were identified across all samples. Alpha diversity did not significantly differ between PEx and end-of-treatment, but it did differ significantly between end-of-treatment and follow-up. Beta diversity, measured using the Bray-Curtis index, showed significant differences between time points, largely driven by the shift in end-of-treatment samples. Only *Gemella haemolysans* was significantly increased in PEx compared to end-of-treatment. *Gemella haemolysans* and *Streptococcus salivarius* were increased in follow-up compared to end-of-treatment. Ten bacterial species (4.5%) showed significant differential gene abundance across clinical states. *Staphylococcus aureus* accounted for 81% of genes more abundant in PEx compared to end-of-treatment, while *Streptococcus salivarius* accounted for 83% of genes more abundant in follow-up. A total of 1,673 genes were found to be significantly differentially abundant across all comparisons. When comparing follow-up to PEx, only *Veillonella atypica* had differentially abundant genes, all eleven belonging to this species. Of 8,653 identified metabolic pathways, 120 were significantly differentially abundant. In PEx vs. end-of-treatment, all 106 pathways were in *Staphylococcus aureus*; in follow-up vs. end-of-treatment, all 66 pathways were in *Streptococcus salivarius*. The only differentially abundant pathway in follow-up vs. PEx was the UDP-N-acetyl-D-glucosamine biosynthesis I pathway in *Veillonella atypica*. There was significant difference in ppFEV1 between PEx and end of treatment (65.1 vs. 80.3, p < 0.001) and between end of treatment and follow-up (80.4 vs. 70.5, p < 0.001).

This study reveals that a small percentage of bacterial species drive the majority of the observed changes in gene abundance across different clinical states in CF patients with PEx. The prominence of *Staphylococcus aureus* during PEx and *Streptococcus salivarius* during follow-up, coupled with the unique metabolic pathway identified in *Veillonella atypica*, suggests specific roles for these species in the disease process. The observed increase in *Staphylococcus aureus* genes associated with persistence and antimicrobial resistance during PEx, and its dominance in uniquely abundant genes, supports its strong association with decreased lung function. The positive correlation between *Veillonella* and milder CF disease reported in previous studies aligns with our finding of increased *Veillonella atypica* metabolic activity during the recovery phase. This metabolic activity may play a crucial role in improving lung function, possibly through its involvement in community interactions and nutrient availability. The results emphasize the importance of considering bacterial metabolic potential in addition to species abundance when studying CF lung disease.

This study demonstrates that bacterial metabolic potential, rather than solely bacterial abundance, provides valuable insights into the dynamic changes across CF clinical states. The involvement of *Staphylococcus aureus*, *Streptococcus salivarius*, and *Veillonella atypica* in the observed changes highlights their potential roles in PEx pathogenesis and recovery. Future research should focus on longitudinal studies to further investigate these metabolic markers as predictors of PEx and to explore the specific mechanisms through which these bacterial species influence the CF lung environment.

This study's limitations include its single-center design, its focus on children and young adults capable of producing sputum, and the collection of data around a single PEx event rather than longitudinally. These limitations may impact the generalizability of findings and hinder a complete understanding of long-term effects. The limited sample size may also reduce the statistical power to detect subtle changes in the microbiome.