Established and Novel Human Translational Models to Advance Cystic Fibrosis Research, Drug Discovery, and Optimize CFTR-Targeting Therapeutics

Abstract

To find a cure for cystic fibrosis, there has been tremendous progress in the development of treatments that target the basic defect in the protein channel, CFTR. However, 10% of cystic fibrosis patients have rare CFTR mutations that are still without an approved CFTR-targeting drug. To identify relevant therapies for these patients, culture models using nasal, bronchial, and rectal tissue from individual patients allow functional, biochemical, and cellular detection of drug-rescued CFTR. Additionally, novel systems such as induced pluripotent stem cell-derived models are utilized to characterize CFTR mutations and identify treatments. State-of-the art translational models were instrumental for CFTR modulator development and may become important for gene-based drug discovery and other novel therapeutic strategies.

Introduction

Cystic fibrosis (CF) is an autosomal recessive disease resulting from mutations in the CF transmembrane conductance regulator (CFTR) gene [1], which encodes an ion channel that transports chloride and bicarbonate, playing an important role in hydration of many epithelial surfaces. After discovery of the CFTR gene in 1989, preclinical research with in vitro human cell models paved the way for the development of CFTR-targeting therapeutics that permit successful treatment of the basic defect in CF. These CFTR modulators are small-molecular compounds known as correctors that augment transfer of mutant CFTR to the apical membrane, and potentiators that increase CFTR channel activity [2].

90% of people with CF (pwCF) in North America, and 80% worldwide carry the F508del CFTR mutation. Unfortunately, 10% of the CF population or more, depending on ethnicity, have rare mutations for which CFTR modulator therapies are not available. To reach these patients and to improve upon available therapies, numerous CFTR-targeting compounds and reagents are currently in the clinical pipeline including novel mRNA- and DNA-based gene therapy therapeutics, read-through reagents for premature stop codons, and advanced modulator compounds (https://www.cff.org/Trials/Pipeline), and thus, personalized models for CF research remain in high demand for predicting drug efficacy for CF individuals.

CF research has developed and employed specific, physiologically relevant human assay systems to advance discovery of CF drugs. Primary human bronchial epithelial (HBE) and nasal epithelial (HNE) cells are typically grown at air-liquid interface (ALI) as planar cultures to study electrophysiological properties that reflect the function of epithelial ion channels such as CFTR [3–7]. Such models are imperative for efficiently identifying and screening compounds before they enter clinical trials, maximizing the likelihood of achieving clinically meaningful improvements in CFTR function, thus facilitating rapid progression of clinical trials toward more effective CF treatments. This review will highlight new and improved models that have been utilized for identifying effective CFTR-targeting therapies for pwCF.

CF Therapeutics

The first FDA-approved CFTR modulator was the potentiator ivacaftor (IVA), which is the active ingredient of a drug that improves the function of the CFTR gating mutant G551D [8,9]. While IVA or the CFTR corrector lumacaftor (LUM) alone did not significantly improve lung function in the CFTR folding mutant, F508del [10], combining LUM with IVA or combining the newer corrector tezacaftor (TEZ) with IVA resulted in modest lung function improvements in clinical trials in pwCF homozygous for F508del CFTR [11–14]. The recently approved triple therapy drug (next-generation corrector elexacaftor (ELX) with TEZ and IVA; ELX/TEZ/IVA) showed substantial efficacy in phase 3 clinical trials [15,16] and was more robust than IVA and dual combination therapies in some populations. The improvement in lung function measured as forced expiratory volume in 1 second percent predicted compared to sex- and age-matched healthy lungs (FEV1pp) in pwCF receiving ELX/TEZ/IVA is significant, with an overall improvement of FEV1pp of at least 10% [16,17], which is comparable to what was observed in G551D pwCF treated with IVA. In addition, improvement in the gastrointestinal (GI) system was observed in patients taking ELX/TEZ/IVA.

Personalized Models

The FDA approval of CFTR modulators for certain CFTR mutations was based on FRT cell line data. Although cell lines are useful to study mechanistic defects of CFTR mutations, they may not constitute a reliable physiological system for predicting all drug effects that are observed in epithelial tissues [18–21]. Some CFTR therapeutics such as read-through reagents prevented premature termination mutations in cell lines but did not work in primary epithelia where extensive nonsense-mediated mRNA decay is observed [22–25]. Identifying therapeutics that rescue CFTR splicing mutations are important for pwCF with such mutations. Studies on splicing mutations were previously conducted in cell lines but are more complicated in vivo and are therefore being intensively pursued [26]. In differentiated epithelia, CFTR localization and function are determined by cell-type specific expression and spatial interactions [27] and highly modulated by environmental factors such as inflammation [28].

Human in vitro models have been crucial for testing CFTR rescue in a relevant physiological environment. As patients respond differently to drugs, the most effective models for predicting clinical responses to therapeutics are cultures derived from tissue samples of individual patients; a personalized medicine approach. These cultures can be used in multiple assays to identify, optimize, and confirm therapies. In addition, it is important to have models available for examination of pharmacokinetics and pharmacodynamics of drugs [29].

Well-established in vitro models are differentiated cultures of HBE and HNE cells on membranes at ALI [30]. ALI culture protocol details (i.e., passage number of cells, differentiation time, type of media, supplements and insert type) may affect cell type composition, the magnitude of ion channel expression and activity, and the amount of CFTR protein that is rescued [31,32]. A common method to assess mutant CFTR rescue in these cultures is to observe electrophysiological responses of CFTR measured as short-circuit currents in Ussing chambers by treatment with forskolin that leads to CFTR activation by cAMP-dependent protein kinase, PKA. Basal CFTR currents before further activation by forskolin may also be indicative of restored CFTR function, as well as a diminution of the epithelial sodium channel (ENaC) activity. Furthermore, activation of CFTR by nucleotides (i.e., UTP or ATP) in the presence of the calcium-activated chloride channel (CaCC) TMEM16A inhibitor offers the possibility to evaluate an additional physiological activation route. Restoration of mucociliary clearance (MCC) of CF HNE and HBE ALI cultures is less commonly used for primary drug screenings; however, they can be investigated to evaluate the ability of a CFTR-targeting drug to restore airway surface liquid homeostasis, normalize mucin concentrations, and ciliary dynamics [33,34], which are all important components of a well-functioning MCC. A recent publication demonstrates the importance of CFTR modulators on airway hydration and mucus properties [33].

Patient specimens from airway and GI tissue can be used to generate organoids for testing rescue of mutant CFTR function [35–37]. Depending on the culture method, airway organoids can be oriented such that the apical membrane faces inward or outward, and rescue of mutant CFTR can be quantitated by changes in organoid size. These methods vary by tissue type and are described in more detail, below. An overview of the most commonly used models to evaluate CFTR therapeutics is shown in Figure 1.

Figure 1: Translational human in vitro models for CF research and drug discovery.

Processing and expansion of nasal bronchial and GI epithelial tissues to form 2D planar and 3D spheroid cultures. Expanded HNE and HBE cells are either seeded on membranes form planar cultures or in matrigel to form spheroids. GI organoids develop directly from partially digested tissues. Planar and spheroid cultures can be utilized to evaluate pharmacological rescue of CFTR rescue by various assays that evaluate CFTR function and maturation. Some image panels were obtained from Servier Medical Art (smart.servier.com).

Bronchial Cultures

Primary HBE cells derived from explant lungs of pwCF and cultured at ALI have been the gold standard for studying the efficacy of CFTR modulators [8,38,39]. In addition, HBE can be obtained from living patients via bronchial brushings, which provide the potential to identify optimal therapeutics for individual patients (personalized medicine). To expand resources available for testing of CFTR modulators, CF HBE cells can be conditionally reprogrammed and maintained to higher passage numbers [32]. Organoids/spheroids from airway epithelial cells are obtained by seeding of CF HBE in matrigel, in which the movement of ions by luminal CFTR channels drives fluid toward the center, resulting in swelling that is not observed in CF patient-derived organoids where CFTR is defective. These models are suitable to quantitate CFTR-targeting therapeutics that induce CFTR-dependent spheroid swelling and are also an appropriate model to study CF pathophysiology [37,40]. In addition, a novel method for generating bronchial organoids with externally oriented apical membranes in mixed matrix components was recently published [41]. Expanded cultures of bronchial epithelial cells can be utilized as an additional source of tissue to create spheroids for more testing of CFTR-targeting therapeutics.

Nasal Cultures

HNE cells have many properties in common with HBE cells and form polarized, pseudostratified epithelia mimicking in vivo airways and the expression of ion channels including CFTR, ENaC, and CaCCs. Similar to HBE cells, HNE cells can be derived by brushing or scraping but the collection is far less invasive than for HBE cells [42]. Thus, patient-derived HNE cultures differentiated at ALI have become a standard model in CFTR modulator testing [42–49]. Epithelial cell types such as ionocytes, ciliated cells, and secretory cells differ between nasal cells and large (bronchi) and small (bronchioles) airways [50–53]; however, HNE cultures appear to recapitulate many of the bioelectric properties of differentiated HBE and respond in a similar fashion to CFTR modulators [42,44,54]. HBE and HNE cultures can be used to evaluate the effects of CFTR modulators on reversing secondary CF phenotypes such as decreased MCC. Similar to HBE, conditionally reprogramming of HNE is available as a method to expand patient culture lifetime without majorly affecting CFTR function [55,56]. CFTR-mediated chloride currents in HNE cells correlated with patients’ sweat chloride concentrations, a common method to detect CF [55]. Thus, HNE cells are recognized as a non-invasive surrogate for HBE cells in many preclinical studies of CFTR modulators [36,42–46,55,57–63]. Patient-derived nasal tissue can be grown in suspension, creating nasospheroids with CFTR channels on the outer surface [36]. Upon rescue of mutant CFTR, ion transport and fluid will move outward toward the media, causing the spheroids to shrink, which can be quantitated [36].

Intestinal Cultures

In 2013, it was shown that organoids derived from patients’ intestines can be utilized to study drug rescue of mutant CFTR function [64]. A recent study showed that intestinal organoids can be used to test for CFTR rescue of nonsense mutations, in which 5 different therapies were combined: the 3 CFTR modulators in ELX/TEZ/IVA plus a compound that induces translational readthrough and a compound that inhibits nonsense mRNA-mediated decay. This is very promising for pwCF with nonsense CFTR mutations that are not eligible for ELX/TEZ/IVA [65]. Another recent study using intestinal organoids showed the importance of testing compounds from different companies to optimize efficacy of F508del CFTR rescue [66]. Although F508del CFTR in patient-derived cultures typically responds well to rescue by triple therapy, cultures from different patients with the same CFTR genotype do not respond in a similar fashion, which may be due to genetic traits other than CFTR that may affect CFTR modulator efficacy and therefore should be further examined [67,68]. Additional models using different intestinal tissues and different culture methods/scaffolds may also have the potential to be used for CF research [69–71].

As clinical improvement of pwCF and in vitro readout by rectal organoids appear to be correlated [65,72,73], large efforts such as HIT-CF Europe aim to expand organoid-based screenings of drugs from multiple companies to pwCF with rare mutations (https://www.hitcf.org/). Furthermore, GI organoids are not only utilized for selection of optimized treatments for rare CFTR mutations but also for evaluation of novel read-through therapies for nonsense mutations and gene therapeutic approaches [74,75]. Additionally, GI organoids can be seeded on membranes to form monolayers that develop to differentiated cultures that are analyzed electrophysiologically in Ussing chambers to study restoration of CFTR-mediated currents in CF planar cultures [76]. Recent publications demonstrate the power of directly comparing clinical data with in vitro data from patient-derived intestinal cultures [77,78]. GI organoids are the most straightforward model to develop from only partially digested tissue specimens. These can be subsequently cultured on inserts as monolayers, offering the following advantages: 1) they can be analyzed in Ussing chambers, and 2) the apical surface is directly accessible for treatments.

Other Gastrointestinal Cultures

CF patients frequently suffer from CF-related diabetes (CFRD), which can lead to glucose imbalance that can in turn, augment the severity of CF disease. It is therefore important to develop a relevant pancreas model for testing potential therapies that improve these glucose imbalances. A pancreas-on-a-chip model was developed using patient-derived pancreatic ductal epithelial cells (PDECs) and pancreatic islets, allowing for examination of the relationship between these cell types [79]. CFTR is expressed in PDECs and inhibition of CFTR channel function leads to a decrease in insulin secretion. To alleviate the effects of CFRD, this pancreas-on-a-chip model can be utilized to test CFTR modulators for their ability to improve glucose imbalances in individual patients. In addition, patient-derived cultures of biliary tissue (cholangiocytes) can also be used to test for rescue of CFTR function [80,81].

Induced Pluripotent Stem Cells

Induced pluripotent stem cells (iPSCs) are adult somatic (e.g., skin or blood) cells that have been reprogrammed, bringing the cells back to an embryonic-like pluripotent state [82]. This allows the creation of an unlimited source of any type of human cell for therapeutic testing. Although creating lung tissue from stem cells is very complicated, in 2015, iPSCs were created from F508del pwCF, corrected with wild-type CFTR gene sequences, and then differentiated into airway epithelial cells [83]. iPSC-derived lung epithelium can be used to generate lung organoids that mimic lung tissue [84–86]. Furthermore, iPSCs can be used to model defects in ciliary function, which can be beneficial for measuring defective MCC in CF cultures [87]. iPSC-derived lung progenitor cells can be set up in a high-throughput platform, allowing studies that measure rescue of mutant CFTR function [88]. In addition, for GI studies, iPSCs were used to create pancreatic duct-like organoids that expressed CFTR [89]. Overall, iPSC technology may be particularly beneficial for pwCF with rare CFTR mutations such as nonsense mutations that do not yet have approved CFTR-targeting therapies.

Preclinical Models in Development

Exciting developments are ongoing using cells not only from various GI segments, but also other tissues such as sweat ducts, submucosal glands, and liver [90–92] for translation modeling of CF. As CF disease is thought to be initiated in small airways, HBE cells from small and large airways are utilized as separate models to display typical characteristics of cell populations found in these specific regions [50]. To study engraftment with HBE cultures expressing wild-type CFTR, a cell therapy approach was developed involving repopulation CF HBE cultures, thereby creating a population of cells that express CFTR with normal function [93]. Advanced models aim to incorporate environmental conditions found in the CF lungs such as infection, inflammation, mucus burden, and hypoxia [28,94]. In testing various drugs for their ability to repair the defects of CFTR mutations, it is important to consider how the efficacy of drugs may be affected by nearby tissues, cells, and secreted reagents such as the endothelium, immune system cells and cytokines, and bacteria. To address this, a CF Airway Chip was created that includes CF bronchial epithelial cells grown at air-liquid interface, lung endothelium, and dynamic fluid flow that can deliver immune cells [95]. These CF Airway Chips accurately represent the human airway in vivo, allowing direct testing of various drugs to rescue rare CFTR mutations. Other organs-on-a-chip are also in development for CF research, including multi-tissue organs-on-a-chip to address the effects of different organs on each other [96,97].

Conclusions

In vitro models derived from tissues of pwCF are vital for the development of effective CFTR therapeutics. Personalized models are important as patients respond differently to treatments. Using tissue that is relevant to CF (bronchial, nasal, and GI epithelial cells) collected from pwCF for drug testing is an effective method for identifying and optimizing therapeutics for each patient. Cultures grown in planar (2D) and spheroid or organoid (3D) formats are used as models of CF disease in multiple drug testing assays. These in vitro translational models have been crucial for understanding CF pathophysiology and CFTR regulation and were instrumental in identifying effective CFTR modulators for pwCF. Newer models such as iPSC-derived cultures and organs-on-a-chip allow the use of additional material and more advanced methods, respectively. In the future, researchers will continue to tackle remaining complicated cellular issues with models that simulate inflammation, infection, mucus, and MCC, and enable gene therapy studies that accurately predict clinical outcomes for pwCF in need of effective therapies.

Highlights.

• Relevant models of CF are critical for predicting clinical outcomes of therapeutics

• Nasal, bronchial, and rectal 2D and 3D cultures have been key for drug discovery

• Personalized medicine is important as individuals respond differently to drugs

• Novel models such as iPSCs and organs-on-a-chip will improve drug testing

• Advanced personalized models will be useful for future gene therapy drug discovery