Poster: 34Bile acids as potential immunomodulatory signals

Abstract

Background

Plasma and sputum from people with CF (pwCF) suppress macrophage functions and it is unknown what induces this suppression. This suppression could be mediated by bile acids (BA), as BA are dysregulated in CF in metabolomic studies of clinical samples and mechanistic studies showing direct CFTR effects on BA regulation. Importantly, some BA species and their derivatives can regulate gut and liver immune cell function by reducing pro-inflammatory cytokine production and phagocytosis. However, it remains unclear whether systemic and airway macrophages respond to BAs, and whether BA dysregulation persists with CFTR modulator therapy.

Plasma and sputum from people with CF (pwCF) suppress macrophage functions and it is unknown what induces this suppression. This suppression could be mediated by bile acids (BA), as BA are dysregulated in CF in metabolomic studies of clinical samples and mechanistic studies showing direct CFTR effects on BA regulation. Importantly, some BA species and their derivatives can regulate gut and liver immune cell function by reducing pro-inflammatory cytokine production and phagocytosis. However, it remains unclear whether systemic and airway macrophages respond to BAs, and whether BA dysregulation persists with CFTR modulator therapy.

Methods

Plasma samples from pwCF pre- and post-elexacaftor/tezacaftor/ivacaftor (ETI) were assessed for BAs and related metabolites. Single-cell RNA sequencing data from the Chan-Zuckerberg CellxGene database were mined to determine if BA receptors are expressed in lung-resident macrophages. Human monocyte-derived macrophages (MDMs) from healthy and CF donors were differentiated with 20% human serum, exposed to BA for 24hrs, and BA effects on phagocytosis and intracellular killing were assessed.

Plasma samples from pwCF pre- and post-elexacaftor/tezacaftor/ivacaftor (ETI) were assessed for BAs and related metabolites. Single-cell RNA sequencing data from the Chan-Zuckerberg CellxGene database were mined to determine if BA receptors are expressed in lung-resident macrophages. Human monocyte-derived macrophages (MDMs) from healthy and CF donors were differentiated with 20% human serum, exposed to BA for 24hrs, and BA effects on phagocytosis and intracellular killing were assessed.

Results

Untargeted metabolomics revealed two BAs in the top 10 most elevated metabolites in CF plasma (vs healthy), and this difference persisted following ETI. We further investigated BA levels using a quantitative panel of 64 BA species. Hyocholic, cholic, and chenodeoxycholic acids were all significantly elevated in plasma from pwCF pre-ETI. This difference persisted with ETI, with a further increase in hyocholic acid. As BA alterations can influence other signaling molecules, such as the production of fatty acids like oxylipins, we also quantified both esterified and non-esterified versions of 50 different oxylipins. Non-esterified oxylipins were not different between CF and healthy plasma, but three esterified oxylipins were significantly elevated in CF plasma. Two in the 5-LOX pathway remained elevated post-ETI, while one related to linoleic acid normalized. Additional human samples (plasma and airway) pre- and post-vanzacaftor/tezacaftor/deutivacaftor are under analysis for untargeted metabolomics, targeted BA and oxylipin quantification, RNA expression, and cytokines.

At the mechanistic level, we explored whether BAs influence macrophage effector functions. We found lung and liver macrophages express membrane and nuclear BA receptors at similar levels. BA treatment did not alter MDM phagocytosis or intracellular killing of P. aeruginosa and S. aureus. Given macrophage plasticity, we then tested if BA receptor expression or BA responses changed with macrophage polarization. M1-polarized (IFNγ & LPS) macrophages upregulated TGR5 RNA, the main membrane BA receptor, but not its surface protein expression. However, BA-treated M2-polarized (dexamethasone) macrophages had a blunted transcriptional response to LPS stimulation. This may suggest a TGR5-independent response. We are currently exploring if BA exposure during differentiation or polarization alters macrophage BA receptor expression and responses.

Untargeted metabolomics revealed two BAs in the top 10 most elevated metabolites in CF plasma (vs healthy), and this difference persisted following ETI. We further investigated BA levels using a quantitative panel of 64 BA species. Hyocholic, cholic, and chenodeoxycholic acids were all significantly elevated in plasma from pwCF pre-ETI. This difference persisted with ETI, with a further increase in hyocholic acid. As BA alterations can influence other signaling molecules, such as the production of fatty acids like oxylipins, we also quantified both esterified and non-esterified versions of 50 different oxylipins. Non-esterified oxylipins were not different between CF and healthy plasma, but three esterified oxylipins were significantly elevated in CF plasma. Two in the 5-LOX pathway remained elevated post-ETI, while one related to linoleic acid normalized. Additional human samples (plasma and airway) pre- and post-vanzacaftor/tezacaftor/deutivacaftor are under analysis for untargeted metabolomics, targeted BA and oxylipin quantification, RNA expression, and cytokines.

At the mechanistic level, we explored whether BAs influence macrophage effector functions. We found lung and liver macrophages express membrane and nuclear BA receptors at similar levels. BA treatment did not alter MDM phagocytosis or intracellular killing of P. aeruginosa and S. aureus. Given macrophage plasticity, we then tested if BA receptor expression or BA responses changed with macrophage polarization. M1-polarized (IFNγ & LPS) macrophages upregulated TGR5 RNA, the main membrane BA receptor, but not its surface protein expression. However, BA-treated M2-polarized (dexamethasone) macrophages had a blunted transcriptional response to LPS stimulation. This may suggest a TGR5-independent response. We are currently exploring if BA exposure during differentiation or polarization alters macrophage BA receptor expression and responses.

Conclusions

scRNA-seq data and in vitro experiments confirmed that lung macrophages express BA receptors at physiologically relevant levels, while metabolomics studies revealed persistent BA alterations in CF. Ongoing studies using alternative differentiation conditions, repeated BA exposure, and human samples will clarify whether BAs contribute to macrophage-mediated bacterial tolerance in CF.

scRNA-seq data and in vitro experiments confirmed that lung macrophages express BA receptors at physiologically relevant levels, while metabolomics studies revealed persistent BA alterations in CF. Ongoing studies using alternative differentiation conditions, repeated BA exposure, and human samples will clarify whether BAs contribute to macrophage-mediated bacterial tolerance in CF.