We describe a human nasal epithelial (HNE) organoid model derived directly from patient samples that is well-differentiated and recapitulates the airway epithelium, including the expression of cilia, mucins, tight junctions, the cystic fibrosis transmembrane conductance regulator (CFTR), and ionocytes. This model requires few cells compared to airway epithelial monolayer cultures, with multiple outcome measurements depending on the application. A novel feature of the model is the predictive capacity of lumen formation, a marker of baseline CFTR function that correlates with short-circuit current activation of CFTR in monolayers and discriminates the cystic fibrosis (CF) phenotype from non-CF. Our HNE organoid model is amenable to automated measurements of forskolin-induced swelling (FIS), which distinguishes levels of CFTR activity. While the apical side is not easily accessible, RNA- and DNA-based therapies intended for systemic administration could be evaluated in vitro, or it could be used as an ex vivo biomarker of successful repair of a mutant gene. In conclusion, this highly differentiated airway epithelial model could serve as a surrogate biomarker to assess correction of the mutant gene in CF or other diseases, recapitulating the phenotypic and genotypic diversity of the population.

We have developed a human bronchial epithelial (HBE) cell and endothelial cell co-cultured microfluidic model to mimic the in vivo human airway. This airway-on-a-chip was designed with a central epithelial channel and two flanking endothelial channels, with a three-dimensional monolayers of cells growing along the four walls of the channel, forming central clear lumens. These cultures mimic airways and microvasculature in vivo. The central channel cells are grown at air-liquid interface and show features of airway differentiation including tight-junction formation, mucus production, and ciliated cells. Combined with novel micro-optical coherence tomography, this chip enables functional imaging of the interior of the lumen, which includes quantitation of cilia motion including beat frequency and mucociliary transport. This airway-on-a chip is a significant step forward in the development of microfluidics models for functional imaging.

In vitro biomarkers to assess cystic fibrosis transmembrane conductance regulator activity are desirable for precision modulator selection and as a tool for clinical trials. Here, we describe an organoid swelling assay derived from human nasal epithelia using commercially available reagents and equipment and an automated imaging process. Cells were collected in nasal brush biopsies, expanded in vitro, and cultured as spherical organoids or as monolayers. Organoids were used in a functional swelling assay with automated measurements and analysis, whereas monolayers were used for short-circuit current measurements to assess ion channel activity. Clinical data were collected from patients on modulators. Relationships between swelling data and short-circuit current, as well as between swelling data and clinical outcome measures, were assessed. The organoid assay measurements correlated with short-circuit current measurements for ion channel activity. The functional organoid assay distinguished individual responses as well as differences between groups. The organoid assay distinguished incremental drug responses to modulator monotherapy with ivacaftor and combination therapy with ivacaftor, tezacaftor, and elexacaftor. The swelling activity paralleled the clinical response. In conclusion, an in vitro biomarker derived from patients’ cells can be used to predict responses to drugs and is likely to be useful as a preclinical tool to aid in the development of novel treatments and as a clinical trial outcome measure for a variety of applications, including gene therapy or editing.

A major unmet need in the cystic fibrosis (CF) therapeutic landscape is the lack of effective treatments for nonsense CFTR mutations, which affect approximately 10% of CF patients. Correction of nonsense CFTR mutations via genomic editing represents a promising therapeutic approach. In this study, we tested whether prime editing, a novel CRISPR-based genomic editing method, can be a potential therapeutic modality to correct nonsense CFTR mutations. We generated iPSCs from a CF patient homozygous for the CFTR W1282X mutation. We demonstrated that prime editing corrected one mutant allele in iPSCs, which effectively restored CFTR function in iPSC-derived airway epithelial cells and organoids. We further demonstrated that prime editing may directly repair mutations in iPSC-derived airway epithelial cells when the prime editing machinery is efficiently delivered by helper-dependent adenovirus (HDAd). Together, our data demonstrated that prime editing may potentially be applied to correct CFTR mutations such as W1282X.

Although major advancements have been made in the therapeutics for people with cystic fibrosis (PwCF), many still require the use of multiple medications to manage acute exacerbations of disease and maintain health. Iterative trial and error processes of pharmacotherapeutic management can be optimized by assessing and incorporating pharmacogenetics. For 82 PwCF, we reviewed 2 years of medication use and response history and interrogated metabolizer status for common pharmacogenes, revealing 3336 medication exposure events (MEEs) to 286 unique medications. As expected, the more frequent MEEs were those commonly used to treat cystic fibrosis (CF), such as antibiotics and respiratory medications. Antibiotics also comprised 56.7% of the undesirable drug responses. The impact of gene variants on drug responses was assessed using Pharmacogenomics Knowledgebase (PharmGKB) and Clinical Pharmacogenetics Implementation Consortium (CPIC) guidelines. Thirty‐three (11.5%) medications have strong evidence of genetic influence on response as evidenced by gene‐based dosing guidelines. 110 (38.5%) unique medications had at least one association with a very important pharmacogene, whereas 143 (50%) were associated with at least one clinical or variant annotation. Over 97% of participants had at least one actionable genotype. Eleven (13.4%) patients with an actionable genotype, taking the impacted medication, had an undesirable drug response described in the CPIC guidelines that could potentially have been mitigated with a priori knowledge of the genotype. PwCF take many medications for disease management, with frequent dose changes to elicit a desired clinical effect. As CF care evolves, implementing pharmacogenetics testing can improve efficiency and safety of prescribing practices using precision selection and dosing at medication initiation.