Azithromycin promotes proliferation, and inhibits inflammation in nasal epithelial cells in primary ciliary dyskinesia – Scientific Reports

The airway epithelium plays an important role as an innate defence mechanism in the airways—it represents a structural barrier, provides removal of inhaled microbes and pollutants by muco-ciliary clearance, and produces anti-microbial peptides and cytokines^{28, 29}. One of the key players in the maintenance of airway epithelial homeostasis are the BCs—a subpopulation of airway progenitor cells that are crucial for both physiological epithelial turnover and post-injury repair³⁰.

In PCD, the accumulation of bacteria in the stagnating mucus leads to chronic inflammation which disrupts the integrity of epithelial barrier. PCD patients suffer from recurrent acute and chronic respiratory infection predominantly caused by bacteria such as Hemophilus influenzae, Streptococcus pneumoniae, Moraxella catarrhalis, Staphylococcus aureus and Pseudomonas aeruginosa³¹.

In this work, we aimed to investigate in vitro the inflammatory environment in airway epithelia in PCD. Our data demonstrates under unstimulated conditions pro-inflammatory cytokine production by both BC and MCE PCD cells when compared to WT cells. Previously published data on PCD inflammatory cytokine environment is limited. A few studies have pointed to an enhanced IL-8/CXCL8 production in peripheral blood monocytes and in sputum from patients with PCD^{13, 14, 16}. We detected increased pro-inflammatory cytokine production in PCD epithelial cells, including chemoattractant factors for neutrophils such as GROA/CXCL1 and CXCL8/IL8^{32, 33}; key pro-inflammatory cytokines—IL1A, IL6 and TNFA³⁴; and molecules that function as growth factors for neutrophils, macrophages, dendritic cells, and fibroblasts (GMCSF, GCSF, MCSF, FLT3L, PDGFAA). Our experiments suggest an underlying pro-inflammatory response in the PCD respiratory epithelium in the absence of an acute bacterial infection.

In addition, we find PCD BCs to have defective proliferative responses as evidenced by a delayed progression towards confluence. Although it could be argued that this could be related in part to their underlying pro-inflammatory phenotype, as it was rescued by AZT for the PCD cells, the lack of a response to the detrimental effects of LPS argues for other possible factors being at play. In other contexts, AZT has been shown to enhance epithelial repair by scratch-wound assay³⁵, and promote barrier integrity³⁴. Further investigation of these effects and potential mechanisms behind the dysfunctional proliferation noted in PCD cells are ongoing.

Recent study by Shoemark and Rubo et al. (2021), who applied topological data analysis approach on a group of almost 400 patients with PCD-related mutations demonstrated that disease symptoms vary from severe in CCDC39, variable in DNAH5 to mild in DNAH11¹⁰. From our findings, we propose that this clinical disease severity can be in part explained by the severity of inflammation being also genotype-associated, as cells with a mutation in CCDC39 exhibited highest mean production of pro-inflammatory cytokines when compared to DNAH5, DNAH11 and WT. However, CCDC39 sample in this study was collected from an adolescent patient, so it is important to consider a possible role of age-dependent factors in the severity of inflammation, such as epigenetics. Still, our findings are in agreement with accumulating evidence for a genotype-determined diversity of disease severity in PCD.

Moreover, stimulation with LPS to mimic the conditions induced by bacterial infection led to further increase in the levels of pro-inflammatory cytokines in PCD cells, demonstrating a heightened response. To the best of our knowledge, no similar experiment assessing cytokine production and proliferation capacity in PCD epithelial cell cultures has been performed to this date. There are a few studies that assessed NO production in HNECs derived from PCD patients in a response to bacterial infection and treatment with pro-inflammatory cytokines, however, with inconclusive results^{37, 38}.

In our work, we showed that AZT might play a crucial role in the maintenance of epithelial barrier integrity and inhibition of chronic inflammation in the respiratory mucosa of PCD patients. AZT is known for its antimicrobial and immunomodulatory properties, ability to inhibit bacterial quorum sensing and biofilm formation^{36, 39}. However, exact mechanisms by which AZT helps decrease the frequency of exacerbations in PCD are not fully understood¹².

To assess the effects of AZT on the PCD airway, we treated cell cultures with 2 different concentrations of AZT. Our data showed that AZT at both doses tested can inhibit production of pro-inflammatory cytokines in PCD epithelium both at baseline and post treatment with LPS, and this irrespective of genotype. This effect was observed in both BCs and ciliated cells. Similarly, AZT-induced reduction of inflammatory cytokines in airway cells from lung allograft recipients was observed by Ling et al.⁴⁰.

In our experiment on ciliated cells, AZT further induced increase in the mean production of immunoregulatory cytokine IL1RA. Besides functioning as a natural antagonist to IL1, IL1RA is reported to be involved in alveolar epithelial repair and elimination of bacteria from lungs⁴¹. Our results indicate that AZT modulates the cytokine production by respiratory epithelial cells and might help control inflammation in PCD.

Furthermore, several studies suggested that AZT might promote maintenance of respiratory epithelial barrier integrity. AZT has been reported to alter processing of tight junctions in respiratory epithelium and hence induce epithelial barrier maintenance in HNECs⁴². Slater et al. reported that AZT promoted re-epithelization in healthy human airway epithelium, while LPS lead to the loss of epithelial barrier integrity³⁶. Correspondingly, our results showed significantly increased proliferation in PCD BCs after treatment with 1 µg/mL AZT when compared to vehicle control. In contrast, co-culture of BCs with LPS resulted in the inhibition of cell proliferation and cell growth that could not be rescued by AZT treatment. We speculate that the protective effect of AZT on the airway epithelium is limited to modulating the inflammatory responses and might be insufficient to enhance regenerative responses following a state of acute infectious exacerbation as modelled by our LPS experiments. Our ongoing work aims to explore the mechanisms at play in the epithelial barrier maintenance under the chronic inflammatory milieu present in PCD.

A limitation of our study is the small number of samples included. Collecting samples from multiple patients for each gene (DNAH5, DNAH11 and CCDC39) and at different ages would allow to verify our findings and better assess the variability in the responses seen, as well as permit separating donor characteristics (e.g. age, gender) from mutation specific effects. However, as PCD is a very rare disease, collecting a larger cohort of patients is challenging. Secondly, acute infection in this study was modelled in vitro using LPS rather than induced by direct exposure of cell culture to bacteria. Introducing bacterial infection might modify the cytokine spectrum produced by epithelial cells as well as the ability of AZT to inhibit inflammation. This forms the basis of additional ongoing studies. Lastly, in vitro models differ from in vivo conditions. However, RNA sequencing previously showed that ALI models are highly representative of the conditions in vivo, as the transcriptomic profiles were similar in 96% of the genes, when comparing cells from ALI culture and cells obtained directly from nasal brushings⁴³.