Patient-Derived Cell Models for Personalized Medicine Approaches in Cystic Fibrosis

Abstract

Cystic fibrosis is caused by mutations in the cystic fibrosis transmembrane conductance regulator (CFTR) channel that perturb anion transport across the epithelia of the airways and other organs. To treat cystic fibrosis, strategies that target mutant CFTR have been developed such as correctors that rescue folding and enhance transfer of CFTR to the apical membrane, and potentiators that increase CFTR channel activity. While there has been tremendous progress in development and approval of CFTR therapeutics for the most common (F508del) and several other CFTR mutations, around 10-20% of people with cystic fibrosis have rare mutations that are still without an effective treatment. In the current decade, there was an impressive evolution of patient-derived cell models for precision medicine. In cystic fibrosis, these models have played a crucial role in characterizing the molecular defects in CFTR mutants and identifying compounds that target these defects. Cells from nasal, bronchial, and rectal epithelia are most suitable to evaluate treatments that target CFTR. In vitro assays using cultures grown at an air-liquid interface or as organoids and spheroids allow the diagnosis of the CFTR defect and assessment of potential treatment strategies. An overview of currently established cell culture models and assays for personalized medicine approaches in cystic fibrosis will be provided in this review. These models allow theratyping of rare CFTR mutations with available modulator compounds to predict clinical efficacy. Besides evaluation of individual personalized responses to CFTR therapeutics, patient-derived culture models are valuable for testing responses to developmental treatments such as novel RNA- and DNA-based therapies.

Introduction

Cystic fibrosis (CF) is a life-threatening autosomal recessive genetic disorder caused by dysfunction in CF transmembrane conductance regulator (CFTR)-mediated transport of Cl⁻ and HCO₃⁻ [1]. To date, more than 2000 CFTR variants have been identified, and many of these cause CF. Based on their effect on the protein function, they are grouped into at least seven classes of mutations characterized by complete absence of mRNA or protein synthesis, protein misfolding, impairment of channel gating or conductance, and diminished amounts of CFTR protein levels or stability [2]. Even though highly effective modulator treatments (HEMT) have been available for the treatment of CF for a decade, capable of correcting the basic defect of the CFTR protein, in Europe about 20% and in the US about 10% of people with CF (PwCF) are not eligible for available modulator therapy. One of the main reasons for this exclusion is that many of these PwCF carry rare or ultra-rare CFTR mutations whose responsiveness to modulators is unknown, having often been excluded from clinical trials due to the reduced number of PwCF carrying these specific mutations. Some PwCF bear mutations that do not produce any CFTR protein (i.e. nonsense mutations, deletions, and insertions). The matching of therapies to specific mutations is called theratyping. However, not all PwCF eligible for modulator treatment may respond to them, due to the action of unknown mutations, mutations in modifier genes, and/or epigenetic modifications. Thus, even individuals with identical CFTR mutations may respond differently depending on their demographics, environmental influences, and individuals’ genetic background. In this context, personalized medicine approaches utilizing advanced models based on patient-derived tissues to study and cure CF with the potential to tailor therapy for a specific patient are needed. 2D and 3D patient-derived cell models (PDCM) are considered the most reliable disease models available since they mimic the in vivo ion channel pathophysiology. These culture models include the genetic background of the PwCF and allow a personalized approach to predict the effect of mutations and therapies and to build a biobank for further analyses. In this review, the state-of-the-art approaches in defining the pathogenicity of CFTR variants and their responsiveness to CFTR modulators are presented. This review will focus mainly on the following PDCM for personalized therapy: nasal and bronchial cells models, and rectal organoids.

CF Therapeutics

Currently available HEMT include correctors that rescue F508del folding and augment transfer to the apical membrane, and potentiators that increase CFTR channel activity, permitting successful treatment of the basic CF protein defect [3]. The first approved (2012) HEMT was the potentiator ivacaftor (VX-770), which improves CFTR function in gating mutants [4,5]. Recent development of a triple combination treatment (Trikafta, Kaftrio) consisting of two corrector compounds, [tezacaftor (VX-661) and elexacaftor (VX-445)] together with ivacaftor resulted in significant improvement of clinical responses [6-8]. Trikafta was initially approved by the U.S. Food and Drug Administration (FDA) in 2019 and Kaftrio by the European Medicines Agency (EMA) in 2020 for the treatment of PwCF who had at least one copy of the F508del mutation. Subsequently, in the US, Trikafta was approved for additional mutations and is now available for 178 mutations (see https://www.cff.org/sites/default/files/2021-10/Trikafta-Approved-Mutations-January-2021.pdf). The development of these HEMT opened the way for the treatment of the basic CF protein defect. Several additional CFTR protein modulators as well as RNA and DNA therapeutics are currently under development in the CF pipeline [9-13]. There are however some concerns about mutations that were included in the list of approved mutations for treatment with Trikafta by the FDA [14], since those considerations were at least partially based on results obtained in FRT cells that do not contain the genomic background of CFTR (only cDNA based constructs that do not include the introns of CFTR). These include c.523A>G that was predicted to produce I175V, but instead causes a splicing defect that leads to a deletion of a portion of the CFTR protein (aa175-193) and R1162L that does not appear to cause CF [14,15]. In light of these concerns, and to test the already available and newly designed CF therapies in a more personalized way, PDCM are now the state-of-the-art models. The roles of three main PDCM for CF research, drug development, and personalized testing are discussed below.

Human Nasal Epithelial Cultures

Human nasal epithelial (HNE) cells are easily collected by a non-invasive brushing procedure of both nostrils. Using a soft interdental or endocervical brush, under direct visualization and without anesthesia, the cells are scraped from the middle part of the inferior turbinate [16]. This anatomical structure of the nasal cavity is characterized by the presence of a pseudostratified airway epithelium that shares most of the anatomical and physiological characteristics of airway epithelium of the remaining part of the upper and lower respiratory tract (Figure 1A). The brushed cells, mainly ciliated, goblet, basal stem cells, and rare ionocytes, recapitulate the distribution of the CFTR protein as cells of bronchial origin [17,18]. Once collected, the HNE cells can be used directly for mRNA analysis, assessing the effect of putative splicing mutation [19] or in general to evaluate the effect of silent or deep intronic mutations on the mRNA maturation and stability. Most importantly, the HNE cells can be expanded and cultured in air-liquid interface (ALI) conditions to obtain a pseudostratified epithelium for functional CFTR evaluation. Indeed, the ALI-cultured HNE cells (derived from PwCF with rare mutations) have been successfully used for the evaluation of the responsiveness to CFTR modulators [17,20-28]. The ALI culture can be achieved in different ways. In one method, the HNE cells are collected and expanded for maximum two passages and then seeded onto the microporous membrane supports until they reach complete confluence in submerged conditions. Then the cells are maintained at ALI and fed with a differentiation medium on the basolateral side only. Usually after 20-30 days in ALI conditions the HNE cultures are highly differentiated and used to evaluate the effect of CFTR modulators by Ussing chamber short-circuit current recording experiments (Figure 1B). As an alternative method, to expand the number of available HNE cells, conditionally reprogrammed cultures (CRC) can be created using a cell culture technique that allows cells to be maintained at higher passage number and thus efficiently enhance the supplies of cells available for measurements that are conducted similarly as with primary HNE cells [27]. Furthermore, patient-derived nasal cells can also be cultured in suspension on plates that prevent attachment, creating spheroids with CFTR channels on the outer surface [29] or alternatively in Matrigel matrix, creating spheroids with luminally oriented CFTR [30]. Upon rescue of mutant CFTR with modulators, ion and fluid transport will cause spheroids to shrink or swell in these models, respectively, which can be quantitated to evaluate CFTR function and the effect of the modulators (Figure 1B).

Figure 1. Human CF models for personalized medicine.

A. Illustrations of cell collection for patient-derived models. B. Cartoon depicting culturing and standard assays for nasal, bronchial, and GI/rectal 2D and 3D patient-derived cell models (PDCM).

Human Bronchial Epithelial Cultures

HBE cells are either derived by a bronchial brush (Figure 1B), which is more invasive than a nasal brush or are harvested from explant lungs of PwCF. It is notable that due to better treatment options for PwCF such as HEMT, the number of CF explant lungs has recently declined [31]. HBE models can support the characterization of novel molecular abnormalities caused by rare mutations of CFTR in individual PwCF [32,33]. These in vitro data can be correlated with in vivo clinical presentation, allowing a more precise physiological summary of the effects of each CFTR mutation on CFTR function and subsequent theratyping of rare CFTR mutations [34].

HBE grown at ALI have served as the gold standard for testing CFTR modulators in US [4,8,35], which has resulted in original approval of HEMT, Kalydeco and Trikafta for treatment of PwCF with G551D and F508del mutations. Similar to HNE cells, HBE cells are cultured at ALI as primary cells [36] or grown as CRC to obtain higher passage numbers for testing (Figure 1B) [37,38]. ALI-cultured HBE cultures have been widely used as a platform for theratyping and to characterize new CFTR mutations of uncertain significance to guide therapy optimization [34]. Assays to evaluate CFTR function range from electrophysiological evaluation by transepithelial short-circuit currents, evaluation of CFTR molecular size that shows proper processing, and other measurements such as airway-surface liquid hydration and mucociliary clearance [39].

Seeding of HBE cells in Matrigel or a similar matrix results in spheroids, in which CFTR activation at the apical membrane (facing the lumen) drives fluid toward the lumen followed by spheroid swelling (Figure 1B), as seen with nasal spheroids. This assay allows quantitation of CFTR therapeutics and in addition spheroids can be utilized to examine CF pathophysiology, including aberrant mucus properties and ciliary beat. [34,40-43]. As an alternative to luminally oriented CFTR, spheroids may develop without Matrigel in media with outside-orientated CFTR, which may be later embedded in a matrix to facilitate analyses [44]. As a result of opposite-directed ion and fluid flow, these spheroids shrink when CFTR is activated.

Advanced human airway organ-on-a-chip models are being developed that incorporate fluid flow and include other cell types besides HBE cells such as lung endothelium, alveolar cells, and immune cells [45-49]. While CF airway chip models may represent the physiology of the human airways in vivo, further development may be necessary before they are broadly accepted as a robust model that allows for drug testing.

Human Gastrointestinal Epithelial Cultures

Adult intestinal stem cells were shown to grow into closed 3D structures with protruding buds and a luminal compartment [50]. These 3D structures named organoids are produced from the stem cells present in the base of crypts isolated from rectal mucosa samples that can be collected by suction or forceps biopsy (Figure 1A). While gastrointestinal (GI) biopsies can be isolated from various parts of the GI tract, rectal tissues are most widely used for personalized testing as the isolation procedure is less invasive. The rectal biopsy procedure is painless and feasible to perform at all ages (even in neonates or infants) in every hospital and there is the possibility of remote analysis in a specialized laboratory since the samples can easily be transported for processing [51]. The crypts isolated from the rectal biopsies can be cultured in extracellular matrix and using a multi-component media that allows fast proliferation of the stem cells present in the crypts as 3D structures; the so-called rectal organoids. As the main channel expressed in these rectal ‘organoids’ is the CFTR protein, a specific assay called ‘Forskolin Induced Swelling’ or FIS assay allows measurement of CFTR function [52] (Figure 1B). Activation of ion transport through the CFTR protein by forskolin induces water movements causing the 3D structures to swell, which is a CFTR-dependent response. Compared to the heterologous cells, which are frequently used as models to study CFTR function, this patient-derived primary cell model has the advantage of faithfully reproducing the subjects’ genetic background. In addition, organoid cultures can be bio-banked and expanded indefinitely and the FIS assays in these organoids allow the prediction of the response to CFTR modulators [53,54] (Figure 1B).

The FIS protocol was fully standardized between three independent laboratories during the HIT-CF European project [55], in order to help fulfill one objective of this project that is to obtain EMA validation of the rectal organoid assays as biomarker of CFTR function in PwCF. If this validation is obtained, this patient-derived CFTR assay may become the most advanced in the regulatory track in Europe [56].

In addition to measuring CFTR function and allowing assessment of CFTR modulators, rectal organoids, can also be used to discriminate between PwCF and non-CF individuals. At baseline (without addition of any compound to enhance CFTR function) the morphology of the rectal organoids depends on whether CFTR is functional or not. Organoids from subjects without CF are rounder (more spherical) and show a liquid-filled lumen due to the activity of CFTR. This can be distinguished from organoids of PwCF that do not have a lumen, presenting a more irregular shape (Figure 2, CF Diagnosis). A novel method to quantify these morphologic differences in the organoids has developed and named ROMA for rectal organoids morphology analysis [57] (Figure 1B). Differences between CF and non-CF rectal organoids have already been reported in the first publication: the usage of the FIS assay to analyze CFTR function [52]. However, with ROMA, it was possible to do a semi-automated analysis to quantify the morphological differences using the fluorescence intensity ratio (IR) after calcein staining and circularity index (CI). Using this assay in a group of 22 organoids from subjects without CF and 167 organoids from PwCF, it was found that these two parameters perfectly differentiated subjects with classic CF versus those without CF [57], thus ROMA may help with accurate diagnosis of CF. Currently, the assay is being tested in borderline CF cases to check the applicability of this assay to guide with diagnosis of more complex cases. Furthermore, ROMA may also be used to predict efficacy of CF therapies.

Figure 2. Cartoon depicting suitability of PDCM.

Discussed cell models are suitable for biobanking, diagnosis, obtaining information on molecular defects (CFTR or modifier genes) and therapeutic testing and drug screening. Created with BioRender.com.

In addition to using the 3D organoid structures to study CFTR function and response to modulators, the rectal organoids can also be cultured as polarized 2D cultures on permeable plastic supports and used to measure transepithelial currents in Ussing chambers [58,59], similarly to HNE and HBE (see above sections and Figure 1B).

Conclusions

Treatments targeting mutant CFTR and enhancing its function have become available over the last decade. Even though the FDA approval of highly effective CFTR modulators was partially built on FRT cell line data, it has become obvious that primary epithelial tissues are more reliable than cell lines to evaluate patients’ responses [14,60-64]. One example of these discrepancies is the mutation G970R, that in FRT was shown to benefit from ivacaftor but in a clinical trial no benefit was observed [31] and a study using organoids showed no rescue with ivacaftor in organoids of patients with the G970R mutation, which was found to induce alternative splicing [65]. Nasal, bronchial, and rectal epithelia have proven most useful in analyzing PwCF CFTR function and responses to therapeutics, and thus can support personalized medicine for PwCF. Each model offers its own advantages (Table 1) and has been utilized broadly to assess personalized responses. This review was based on the work presented by the authors in the ECFS Basic Science Conference 2022 and thus did not focus on CF epithelial models derived from human induced pluripotent stem cells (iPSC) or on organ-on-a-chip models that may have great potential for further CF disease modeling and personalized testing.

Table 1.

Most utilized PDCM for personalized medicine for cystic fibrosis.

Model	Source	Key Feature	Culture	Readout	Application
HNE	Nasal Brush or Scrape	Non-invasive isolation procedure. Represents main features of the airway epithelium. Permits testing of responses to CFTR therapeutics.	ALI culture	Isc, Genetic analysis, Theratyping	Common
			Spheroid	Swelling or shrinking, Therapyping	Less common, but more and more established
HBE	Bronchial Brush, Explant Lungs	Invasive isolation procedure. Represents airway epithelium. Well-established assays for testing of responses to CFTR therapeutics.	ALI culture	Isc, Genetic analysis, Theratyping	Very common
			Spheroid	Swelling or shrinking, Therapyping	Less common
Human GI Epithelium	GI/Rectal Biopsy	Represents GI tract. Increasingly used for personalized testing of responses to CFTR therapeutics. Organoid assays have HTS capability. High Levels of CFTR protein in rectal biopsy tissues facilitate testing of efficacy of CFTR therapeutics. ICM with tissues requires instant analysis without excessive storage times (Isc measurements of polarized 2D cultures on permeable supports may overcome this limitation).	Tissue	ICM	Less widespread but accepted
			Planar culture on permeable support	Isc, Genetic analysis	Less common
			Organoid	Swelling, ROMA Genetic analysis, Therapyping	Very common in Europe

Indicated are the main source, key feature, culture model, most common readouts, and the prevalence of application of 2D and 3D patient-derived cell models (PDCM). Short-circuit current (Isc), high-throughput screening (HTS), intestinal current measurement (ICM), rectal organoid morphology analysis (ROMA).

Mechanistic insights into modulator actions and their improvements will not only be applicable for CFTR-targeting treatments but may also be relevant to enhance newly developed RNA and gene therapies and read-though therapeutics that may be utilized in combination with CFTR modulators. Thus, the presented models and assays may have immediate translational relevance in providing methods and knowledge for accurate elucidation of CF drug effects, which will allow optimal therapies for all PwCF.

Overall, PDCM like HNE, HBE, and rectal organoids support assessing in vitro drug efficacy to potentially predict in vivo therapeutic responsiveness and therefore are paving the way to personalized medicine for PwCF. Furthermore, these described models may also be applicable to identify treatment for other human diseases where CFTR function is impaired [66,67].

Highlights.

• Primary human epithelial tissues are more reliable than cell lines to evaluate cystic fibrosis therapies.

• Nasal, bronchial, and rectal epithelia have proven most useful in analyzing CFTR function and responses to cystic fibrosis therapeutics.

• 2D and 3D patient-derived cell models are considered the most reliable disease models since they mimic the in vivo ion channel pathophysiology.

• Personalized medicine approaches utilize advanced models based on patient-derived tissues to tailor therapies for specific patients to cure cystic fibrosis.