Immortalisation of human airway epithelial cells for genome editing applications

Abstract

Aim

In vitro air–liquid interface (ALI) cultures of human airway epithelial cells are widely used to study respiratory biology and disease, as they recapitulate key features of the pseudostratified airway epithelium and reduce reliance on animal models and repeated patient biopsies. However, the limited proliferative capacity of primary airway basal cells restricts long-term studies and genome editing. This study aimed to generate stable, immortalised human airway basal cell lines that retain differentiation capacity and support genetic manipulation.

Methods

Human airway epithelial basal cells from bronchial and nasal origin were immortalised by lentiviral expression of human telomerase reverse transcriptase (hTERT) followed by polyclonal expansion. A panel of cell lines derived from three independent nasal donors was established. Differentiation capacity was assessed in ALI cultures and compared with primary airway epithelial cells. Suitability for genetic manipulation was evaluated using lentiviral overexpression and CRISPR/Cas9-mediated gene knockout and knock-in strategies.

Results

All immortalised cell lines displayed sustained proliferative capacity and efficiently differentiated into major airway epithelial cell types, including multiciliated and secretory cells. Differentiation patterns and functional characteristics closely resembled those of primary airway epithelial cultures. The cell lines supported efficient lentiviral transduction and CRISPR/Cas9-based genome editing, enabling stable gene overexpression, targeted gene disruption, and precise genomic modification.

Conclusions

hTERT-immortalised human airway basal cell lines preserve key features of primary airway epithelium while enabling long-term culture and versatile genome editing. This resource provides a robust platform for airway biology research, disease modelling, and therapeutic development.

Short abstract

This study describes immortalised human airway basal cell lines that retain key epithelial properties, enable long-term expansion and genome editing, and provide a robust platform for airway biology, disease modelling and therapeutic development https://bit.ly/4x3n9oV

Introduction

The human airway epithelium lines the respiratory tract, providing protection against pathogens and particles, while supporting efficient gas exchange in the lower airways [1]. This epithelium is composed of several cell types, including basal cells that act as the proliferating and regenerating population of the epithelial sheet, goblet cells that produce mucus, and multiciliated cells that beat in a coordinated manner to remove mucus deposited by goblet cells from the distal to proximal airways [1]. For decades, air–liquid interface (ALI) culture models have been widely used to study differentiated primary human airway epithelial cells in vitro [2, 3]. This is especially useful for studying epithelial defects in monogenic diseases, such as cystic fibrosis and primary ciliary dyskinesia (PCD) [4, 5]. However, in contrast to commonly used cell lines, primary basal airway cells have limited proliferative capacity, and long-term airway culture is impossible [6–10]. Furthermore, obtaining patient material is both time-consuming and costly, does not provide a scalable source of biological tissue and is subject to significant donor-to-donor variation, which can affect reproducibility and consistency in experimental outcomes. This limits the development of therapeutic strategies and studies of protein function using genetic modifications, because genome manipulations often require multiple rounds of selection and thus long-term culturing [10].

To address this problem, various strategies have been used to extend the time that airway epithelial cells can be maintained in culture. One approach involves optimisation of culture conditions or use of co-culture systems [11, 12]. In addition, genetic approaches have been employed to overcome replicative senescence, typically by introducing oncoproteins, such as the simian virus 40 large T antigen (SV40), human papillomavirus (HPV) E6 and E7 proteins or Bmi-1, or by manipulating pathways related to cell cycle regulation, for example, by expressing cyclin-dependent kinase (Cdk)4 or by expressing human telomerase reverse transcriptase (hTERT), either alone or in combination with oncogenes [3, 6, 9, 13–22]. These immortalised airway cell lines have an extended lifespan and, to various degrees, retain characteristics of differentiated airway epithelium. For instance, although 16HBE140 and BEAS-2B can be used in studies of drug absorption and transport, they do not form polarised epithelial sheets or differentiate into multiciliated cells [15, 16]. Other cell lines, such as VA10, BCi-NS1.1, huAEC and UNCN1T, support multiciliated cell differentiation, but often with low proportions of ciliated cells or reduced ciliary beat frequency compared to primary cells [11, 18–23]. Consequently, extending the repertoire of immortalised lines is needed to exclude potential constraints associated with the use of cells that all share the same donor origin and to support the ongoing advances in respiratory disease research, gene-targeted therapies, and personalised medicine.

In the current study, we immortalised human airway epithelial basal cells using lentiviral expression of hTERT. We generated a panel of immortalised cell lines from bronchial and nasal origin and compared three nasal donors. The resulting cell lines retained the ability to differentiate into the major airway epithelial cell types and showed functional characteristics resembling those of healthy primary tissue material. Utilising these immortalised cell lines, we showcased three genome editing strategies: lentiviral protein overexpression, gene knockout and knock-in. We expect that these immortalised cell lines in combination with the genome editing toolbox will help to extend our basic knowledge of airway processes and accelerate testing of new therapies.

Methods

Isolation, expansion and biobanking of primary airway epithelial cells

All individuals included in this study provided broad informed consent for research use of their samples. The study was reviewed and approved by the biobank research ethics committee (TCbio) of the University Medical Center Utrecht (protocol identifiers 16/586 and 22U-0079). Collection, storage, and distribution of airway epithelial materials were coordinated through the Utrecht Platform for Organoid Technology biobank, ensuring standardised procedures, ethical oversight and availability as resource material. Individuals' characteristics (year of birth, sex and relevant clinical history) are listed in supplementary table S1.

Primary nasal epithelial cells (HNECs) were isolated from healthy or PCD individuals through nasal brushing samples and cultured in nasal cell isolation medium (supplementary table S1). Human bronchial epithelial cells (HBECs) were isolated from nontumourous resected airway tissue (HBEC1) or from bronchial biopsies of a patient with nonpulmonary cancer (HBEC2) using bronchial cell isolation medium. Bronchial cell isolation, expansion and differentiation were performed as described previously, with cells maintained at 37°C and 5% carbon dioxide (culture details are provided in the supplementary methods) [24].

Cell proliferation was assessed by population doublings at each passage, calculated as log(Nf)−log(Ni)/log(2), where Nf is the final number of cells after expansion of each subculture and Ni is the initial plated number.

Lentiviral production and infection

Lentiviruses were produced by co-transfection of envelope vector pMD2.9 and packaging vector psPAX2 using Polyethylenimine “Max” (linear, MW 25 000) (Polysciences, PA, USA; #24765). Supernatant of the packaging cells was harvested at 48 h and 72 h post-transfection, filtered (0.45 μm), and incubated overnight at 4°C with a 400 g·L⁻¹ polyethylene glycol 6000-based virus precipitation buffer (ThermoScientific, MA, USA; A17541 0B). The virus suspension was then centrifuged (1500×g, 30 min, 4°C), and the lentivirus pellet was resuspended in PBS.

Basal cells were seeded to achieve 50% confluency in a 12-well plate at the day of infection. 1 h before infection, cells were incubated with 5 μg·mL⁻¹ Polybrene (Millipore, MA, USA). 2.5 μL lentivirus was added dropwise. Following infection with hTERT-containing lentivirus (supplementary table S1), positive cells were selected using 100 μg·mL⁻¹ Hygromycin (Hygromycin B Gold; InvivoGen, CA, USA) until control cells died.

Quantification of cell types

Airway cell types were quantified on immunofluorescence stainings for basal cells (P63) and club cells (CC16) or goblet (mucin 5AC (MUC5AC)) and multiciliated cells (β4 tubulin) together (immunofluorescence protocol is provided in the supplementary methods). Cell numbers were automatically determined and divided over the total cells (4′,6-diamidino-2-phenylindole). Details of quantification are provided in the supplementary methods.

Per cell line and passage (early/late), one or two fields of view (FOV) per replicate (n=3) were analysed. FOV areas were 0.207, 0.016, 0.180 and 0.180 mm² for basal, club, goblet and multiciliated cells, respectively.

Quantitative real-time PCR

Secretoglobin Family 1A Member 1 (SCGB1A1) RNA levels (club cells) were assessed using quantitative (q)PCR, using SYBR Green reaction on a CFX-384 real-time detection system (Bio-Rad, USA) and cell-type-specific primers (supplementary methods).

Transepithelial electrical resistance

The integrity and permeability of the differentiated airway epithelium was assessed weekly during differentiation using transepithelial electrical resistance (TEER) (supplementary methods).

Forskolin-induced swelling assay

Baseline cystic fibrosis transmembrane conductance regulator (CFTR) function was measured using the forskolin-induced swelling (FIS) assay following established protocols (supplementary methods) [24, 25].

High-speed video microscopy and ciliary beat frequency

Ciliary movement in ALI cultures was analysed by measuring the ciliary beat frequency (CBF). ALI-differentiated HNECs were imaged on the Zeiss Cell Discoverer 7 microscope in brightfield mode. The recorded videos were analysed using FIJI Correlesence v.0.0.6 plugin (“Temporal ICS” command) (https://github.com/ekatrukha/Correlescence), resulting in a frequency heatmap for each position from which the CBF (Hz) and beating area (%) were determined (supplementary methods).

Fluorescence-activated cell sorting enrichment of HNECs

Fluorescence-activated cell sorting (FACS) was used to isolate H2B-mNeonGreen and End Binding protein 3 (EB3)-tdTomato positive cells. Basal cells were trypsinised, collected in the wash buffer and centrifuged at 400×g for 5 min (supplementary table S1), then resuspended in 0.5% fetal bovine serum in PBS at ≤20.0×10⁶ cells·mL⁻¹. Sorting was performed on a BD FACS Aria Fusion Sorter (BD Bioscience, CA, USA) (for specifications refer to the supplementary methods).

Lattice light-sheet live-cell imaging

To acquire a single ciliary beat of multiciliated airway epithelial cells with high temporal and spatial resolution, a quarter filter piece was inverted and secured into a 35 mm Attofluor Cell Chamber (Invitrogen, MA, USA) with a plasma-cleaned 25 mm coverslip (Harrick Plasma Cleaner PDC-002) (mounting procedure is described in the supplementary methods) and imaged on a Zeiss Lattice Lightsheet 7 microscope (specifications are provided in the supplementary methods).

CRISPR-Cas9 genome editing

Generation of CRISPR-Cas9 cell lines was achieved by electroporation of immortalised cells with ribonucleoprotein (RNP) complex or two DNA vectors [26]. For Coiled-coil domain 39 (CCDC39) knockout, RNP complexes were prepared by mixing recombinant 2NLS-Cas9 nuclease (20 μM, 2.5 μL), multiguide RNA (mix of three guide RNAs; supplementary table S1) (30 μM, 8.3 μL) (chemically synthesised by Synthego) against CCDC39 and 14.2 μL OptiMEM (Gibco, MA, USA) supplemented with 10 μM Y-27632 (Selleck Chemicals, TX, USA) and incubated at room temperature for 10 min. Next, 1.0×10⁶ cells in 75 μL optiMEM (with 10 μM Y-27632) were added to the RNP complexes and electroporated using the NEPA21 system following established protocols [27]. Cells were then plated in six-well plates containing nasal or bronchial medium.

For H2B-mNeonGreen (H2B-mNG) knock-in, 2.5 μg vector with histone H2B sgRNA and Cas9 and 7.5 μg vector with mNeonGreen flanked by histone H2B homology arms (gift from H. Snippert, Princes Maxima Center for Pediatric Oncology, Utrecht, the Netherlands) (plasmid details are provided in supplementary table S1) were pre-mixed and stored on ice. 0.5×10⁶ basal cells were trypsinised, resuspended in Neurobasal medium (Gibco, MA, USA) supplemented with 1% B-27 supplement, 0.5 mM glutamine, 15.6 μM glutamate and 1% penicillin/streptomycin and combined with pre-mixed DNA vectors (and protein) before electroporation (10 μL electroporation tips, two pulses, 20 ms, 1500 V, Neon system) (InvitrogenMPK5000, MA, USA). Cells were immediately plated in 12-well plates and cultured 24 h without antibiotics to promote cell survival. Interference of CRISPR Edits (ICE) analysis details are provided in the supplementary methods.

Western blotting

CCDC39 protein levels were assessed using Western blotting, using lysates of differentiated cells, which were separated on 7.5% polyacrylamide gel, transferred to polyvinylidene difluoride membrane (Immobilon FL; Sigma-Aldrich, MO, USA) and stained with primary and secondary antibodies followed by chemiluminescence detection (supplementary methods).

Statistical analysis

Statistical analyses (figures 2 and 4) were performed in Prism 8.4.3. using an unpaired, parametric t-test with Welch's correction. Significance was defined as p<0.005 and p<0.01.

FIGURE 2.

Immortalised airway cells retain the ability to differentiate into major airway cell types. a) Immunofluorescence images of primary and immortalised (i) airway cells differentiated at early (e) and late (l) passages and stained for basal cells (P63), goblet cells (MUC5AC), multiciliated cells (β4 tubulin), nuclei (4′,6-diamidino-2-phenylindole (DAPI)) and F-actin (phalloidin). b–e) Percentage of b) basal cells, c) multiciliated cells, d) goblet cells and e) club cells in primary and immortalised airway cells at early (e) and late (l) passages. b) and e) n=3 replicates with one field of view (FOV) per replicate; c) and d) n=3 with two FOV. Statistical tests were performed using an unpaired parametric t-test with Welch's correction. Data are shown as mean±sd, and each point represents a replicate. f) Immunofluorescence images of primary human nasal epithelial cells (HNEC3) differentiated at an early passage and stained for club cells (CC16) and F-actin (phalloidin). g) mRNA expression levels to detect club cells (SCGB1A1) normalised for the geometric mean expression of the housekeeping genes ATP5B and RPL13A in indicated primary and immortalised airway cells differentiated at early (e) and late (l) passages. n=3 for all conditions. Statistical tests were performed using an unpaired nonparametric Mann–Whitney t-test. Data are presented as mean±sd, and each point represents a replicate. ns: nonsignificant. **: p<0.01; ^#: p<0.005.

FIGURE 4.

Lentiviral overexpression of EB3-tdTomato in immortalised primary human nasal epithelial cells (HNEC3i). a) Schematic overview of the workflow to generate HNEC3i expressing EB3-tdTomato. HNEC3i basal cells were infected with lentivirus containing EB3-tdTomato expression cassette; positive cells were enriched using fluorescence-activated cell sorting (FACS) and expanded. HNEC3i EB3-tdTomato cells were differentiated in air–liquid interface culture and used for imaging. b) Immunofluorescence image of HNEC3i basal cells expressing EB3-tdTomato (magenta) and β4B tubulin (green). Box indicates region of zoom. c) Immunofluorescence image of differentiated HNEC3i expressing EB3-tdTomato (magenta) and β4B tubulin (green). Boxes indicate zoomed regions: box 1 shows a nonciliated cell with high EB3-tdTomato expression levels at low and high brightness and contrast setting; box 2 shows a ciliated cell with low EB3-tdTomato expression levels at low and high brightness and contrast setting. d) Immunofluorescence image of cell shown in c) box 2 in the cilia plane to illustrate the localisation of EB3-tdTomato at ciliary tips. e) Flow cytometry analysis of HNEC3i EB3-tdTomato. Cell selection based on TdTomato signal intensity (y-axis) and green autofluorescence was used to improve cell separation (x-axis). Cells with low and high EB3-tdTomato expression levels are indicated. f) Video frames of multiciliated cell visualised with live dye SiR-tubulin (green) and EB3-tdTomato (magenta) during one beat cycle. DAPI: 4′,6-diamidino-2-phenylindole

Results

Immortalised airway basal cells exceed proliferation limit of primary cells and retain differentiation capacity

Primary cells were collected from human bronchial tissues and nasal brushings, and optimised culture conditions were used to selectively proliferate basal cells [25]. Immortalised cell lines were generated by lentiviral expression of hTERT. Basal cells positive for integration were selected using hygromycin (figure 1a). Although serial dilution enabled isolation of individual clones in the past [21, 22], cellular stress responses might be triggered due to low cell density. Here, we reasoned that bulk expansion of basal cells would naturally steer the population towards selection of the healthiest proliferating cells without clonal selection. To assess the variation and reproducibility of the immortalisation strategy with hTERT alone, we cultured primary human epithelial cells obtained from two bronchial regions (HBEC1 and HBEC2) and from three nasal donors (HNEC1, HNEC2 and HNEC3), in parallel. While proliferation of primary HBECs and HNEC3 stalled at approximately 40, 60 and 46 population doublings, respectively (figure 1b), cells from the same individuals expressing hTERT exceeded 78, 112 and 121 population doublings, respectively, and could be cultured for >300 days (figure 1c). Although the doubling time varied slightly from donor to donor, in all cell lines it remained constant over time. This demonstrates that hTERT overexpression alone is sufficient to generate immortalised bronchial cells (HBEC1i, HBEC2i) and nasal cells (HNEC1i, HNEC2i, HNEC3i), which exceed the proliferation limit of primary airway epithelium basal cells.

FIGURE 1.

Human telomerase reverse transcriptase (hTERT) immortalised airway epithelial cells exceed proliferation limit. a) Schematic overview of the workflow to generate immortalised airway cells. Primary basal cells were infected by lentiviral hTERT expression cassette; positive cells were selected using hygromycin, expanded and differentiated in an air–liquid interface (ALI) culture system. b) Growth of primary human bronchial epithelial cells (HBEC1 and HBEC2) and nasal airway (HNEC3) epithelial basal cells. Cell population doublings were plotted against time. Each point represents one passage (p) and the last passage number is indicated for each donor. The enlarged points represent early passages (p4–9) which were used for assays in figure 2. c) Growth of immortalised (hTERT⁺) bronchial (HBEC1i and HBEC2i) and nasal airway (HNEC1i, HNEC2i, HNEC3i) epithelial basal cells. Cell population doublings were plotted against days in culture. Each point represents one passage, and the last passage number is indicated for each donor. The enlarged points represent early (p10–14) and late passages (p23–26), which were used for the assays in figure 2.

To further characterise the immortalised cell lines, we compared their differentiation capacity at early (p10–14) and late (p23–26) passages to primary cells, viable only at early passages (p4–9) (HNEC3, HBEC1 and HBEC2) using immunofluorescence staining with antibodies against the major cell types present in the differentiated airway epithelium (figure 2a). Basal cells (P63) were present in all immortalised cell lines at both an early and late passage (figure 2a, b). Next, we assessed the presence of differentiated cell types (figure 2a–g). All primary cells and immortalised cell lines differentiated into β4 tubulin-positive multiciliated cells and MUC5AC-positive goblet cells (figure 2a, c, d). Immunostaining for CC16 and qPCR for SCGB1A1, both club cell markers, showed that these cells were rare or absent in all our primary and immortalised cell cultures, in line with their low numbers in the upper airway (figure 2e–g) [28]. Because of our interest in multiciliated cell function, the culture medium was supplemented with the Notch inhibitor, DAPT, which promotes differentiation towards multiciliated cells at the expense of secretory cells. This expected shift in cell composition was clearly reflected in the cultures [29].

Although all cultures contained the four major airway epithelial cell types, the relative proportions of these populations varied between donors and across passages. However, no consistent trend was observed, suggesting that these differences do not reflect donor-dependent variation, but are more likely due to minor differences in culture conditions, as also indicated by variation between replicates. Together, these findings indicate that both bronchial and nasal cell lines immortalised by hTERT overexpression maintain the capacity to differentiate into mucociliary epithelium under optimised conditions.

Immortalised airway cells retain functional characteristics of the differentiated epithelium

Next, we assessed several characteristic functions of airway epithelial cells differentiated from immortalised basal cells at late passage numbers (p24–26). First, CBF (Hz) was assessed by imaging cells using high-speed video microscopy. CBF of multiciliated cells differentiated from immortalised cell lines was between 6.5 Hz and 15.5 Hz and slightly lower than that of primary cells (between 7.7 Hz and 20.3 Hz), but remained constant throughout multiple days in ALI culture (figure 3a–c). Second, the barrier function of epithelial cells, which relies on the junctional integrity between cells, was measured by performing TEER assays at different time points during ALI culture differentiation (figure 3d, e). Primary cells stabilised at ∼1000 Ω per cm² at ALI day 7, whereas in immortalised cell lines differentiated at late passages (p24–p26), TEER reached ∼2000 Ω·cm⁻² at ALI day 7 or 14. Last, an important function of airway epithelium is mucociliary clearance. In secretory cells, CFTR protein modulates the transport of anions and fluid in and out of cells to control mucus consistency [30]. A FIS assay can be used to examine CFTR efficiency and ion transport function [24, 31]. We generated airway cystic organoids from one bronchial and one nasal immortalised cell line, HBEC1i and HNEC3i, respectively, at late passage number (p24 and p26) (figure 3f). FIS assay showed that organoids derived from both immortalised cell lines responded to forskolin by swelling approximately two-fold in 1 h, comparable to that which has been published (figure 3f, g) [24]. This confirmed that the immortalised cell lines can be used to generate organoids suitable for measuring CFTR function with FIS assays.

FIGURE 3.

Functional assays with differentiated immortalised airway epithelial cells. a) Legend indicating the colours used to depict data for primary donor cells or immortalised cell lines in parts b–e). b, c) Ciliary beat frequency (CBF) (Hz) of primary (p4–9) and immortalised airway epithelial cells differentiated at late passage number (p24–26). CBF was measured at several time points of differentiation. In all conditions, n=3 replicates. Data are presented as mean±sd, and each point represents a replicate. d, e) Transepithelial electrical resistance (TEER) assay of primary (p4–9) and immortalised airway epithelial cells differentiated late passage number (p24–26), respectively. In all conditions, n=3 replicates. Data are presented as mean±sd, and each point represents a replicate. f) Representative images of forskolin (fsk)-induced swelling determined with calcein green AM ester-stained immortalised primary nasal epithelial cells (HNEC3i) and immortal human bronchial epithelial cell (HBEC1i) organoids, with images taken before the addition of 5 μM forskolin (0 min) and after stimulation (60 min). Insets show zooms of a few organoids. Scale bar (inset)=250 µm. g) Quantification of area under the curve (AUC) of HNEC3i or HBEC1i organoid swelling after 60 min in control conditions and after the addition of 5 μM forskolin. In all conditions, n=2 replicates. Data are shown as mean±sd, and each point represents a replicate.

Together, these assays demonstrate that immortalisation does not affect airway epithelial functions such as ciliary beating, junctional integrity and ion and fluid secretion after long-term culturing. Therefore, they are suitable for modelling diseases where ciliary activity or anion-dependent fluid secretion are affected.

Lentiviral gene expression in immortalised HNECs

Having established immortalised airway epithelial cultures, we next used them for genetic modification. We first set out to visualise microtubules using overexpression of the well-characterised, fluorescently tagged β4B tubulin (encoded by TUBB4B, and recognised by β4 tubulin antibody) and microtubule-binding protein End Binding protein 3 (EB3) [32]. Previous work on epithelial cells from mouse trachea and human lung tissue showed that lentiviral transduction is an effective strategy to express transgenes [9, 33]. To this end, we infected HNEC3i basal cells with a lentiviral construct to express β4B tubulin with a N-terminal green fluorescent protein (GFP) tag and EB3 with a C-terminal tdTomato tag (figure 4a). Confocal microscopy of fixed basal cells showed that GFP-β4B tubulin and EB3-tdTomato decorated microtubules and EB3-tdTomato was enriched at microtubule plus ends as expected (figure 4b). Furthermore, differentiation of this cell line confirmed that EB3-tdTomato specifically localised to ciliary tips, as described previously (figure 4c, d) [34]. This localisation pattern was only visible in a subset of multiciliated cells due to the heterogeneity in EB3-tdTomato expression levels, probably caused by the variation in the number of lentiviral transgene inserts per cell. This is a problem, especially when the localisation of a protein of interest strongly depends on its expression levels, such as with EB3, which binds to microtubule tips at low expression but decorates the whole microtubules at higher expression levels (figure 4c). To overcome this heterogeneity, we selected basal cells with low (4.8% of total cells) and high (0.7% of total cells) EB3-tdTomato expression using fluorescence-activated cell sorting (FACS) of transduced basal cells (figure 4e). Live-cell imaging of differentiated HNEC3i expressing low levels of EB3-tdTomato with a lattice light-sheet microscope showed that EB3 specifically localises to the tips of beating cilia in multiciliated cells (figure 4f). These optimisation steps clearly illustrate the benefit of working with immortalised cells. Therefore, we conclude that lentiviral transduction is an effective approach to overexpress tagged proteins of interest for imaging of fixed and live airway epithelial cells generated from immortalised basal cells.

CRISPR genome-editing in immortalised HNECs: successful knockout and knock-in

To extend our genome editing toolbox, we explored the possibility of CRISPR-Cas9-mediated genome editing of immortalised HNECs. Previously, lentiviral transduction or electroporation were used to introduce the CRISPR-Cas9 machinery into airway epithelial basal cells [35, 36]. In the former study, a single lentiviral vector was used to express single guide RNA (sgRNA), Cas9 and a resistance gene to knockout the gene expressing a transmembrane glycoprotein, Mucin 18 (MUC18), used as a marker for cancer progression [36]. In the latter study, a Cas9-sgRNA RNP complex was used together with a repair template divided over two adeno-associated viruses (AAVs) [35]. Here, we employed a CRISPR-Cas9 editing approach which surpasses the need for lentivirus, AAVs and multiple repair templates [26]. We focused on CCDC39, a protein required for the assembly of inner dynein arms and ciliary motility, the disruption of which is a major cause of PCD [37].

To generate CCDC39 knockout cells, we electroporated three chemically synthesised sgRNAs targeting CCDC39 into HNEC1i, HNEC3i and HBEC1i basal cells in complex with Cas9 protein (figure 5a, b). Introduction of a Cas9/sgRNA complex has been shown to improve editing efficiency [38]. Protein levels of CCDC39 in differentiated epithelium were analysed using Western blotting (figure 5c). Comparison of CCDC39 signal to PCD patient material (HNEC_PCD1 and HNEC_PCD2) confirmed a successful knockout of CCDC39 in HNEC1i and HNEC3i and a strong decrease in protein levels in HBEC1i basal cells (figure 5c). Genomic sequencing and ICE analysis showed that the knockout efficiency was 97%, 73% and 53% for HNEC1i, HNEC3i and HBEC1i, respectively. Those percentages are in line with the Western blot results, which show that HBEC1i is a heterogenous population with only partial knockout. Next, we analysed ciliary beating in ALI differentiated knockout cell lines. Based on the live SiR-tubulin staining, the area occupied by multiciliated cells did not change upon CCDC39 knockout, but the percentage of multiciliated cells with beating cilia decreased compared to the wild type (figure 5d, e), whereas the CBF was unchanged (figure 5f). This confirms that electroporation of a sgRNA-Cas9 complex into immortalised cell lines is an effective strategy to generate knockouts, and that CCDC39 knockout recapitulates the phenotype observed in PCD patients [37].

FIGURE 5.

CRISPR-Cas9 knockouts of CCDC39 recapitulate primary ciliary dyskinesia (PCD) disease phenotype. a) Schematic overview of CRISPR-Cas9 knockout strategy. Various immortalised airway basal cells were electroporated with sgRNA-Cas9 complex, differentiated in air–liquid interface (ALI) culture, validated using Western blot and used for ciliary beat frequency (CBF) assay. b) Scheme of CCDC39 and location of three sgRNAs. c) Western blot illustrating the CCDC39 expression in PCD patient material (primary human nasal epithelial cells (HNEC) PCD1, HNEC PCD2) and wild-type (WT) and knockout (KO) immortalised airway epithelial cells (HNEC1i, HNEC3i and immortalised human bronchial epithelial cells (HBEC1i). Arrows indicate CCDC39 protein. d) Video frames of WT and CCDC39 KO cells HNEC1i, HNEC3i and HBEC1i incubated with live dye SiR-tubulin with corresponding CBF maps. e, f) Area of multiciliated cells with e) beating cilia and f) average CBF in Hz in WT and CCDC39 KO cells HNEC1i, HNEC3i and HBEC1i. In all conditions, n=2 replicates. Data are presented as mean±sd, and each point represents a replicate. hTERT: human telomerase reverse transcriptase.

To test whether immortalised airway epithelial cells can also be used to obtain CRISPR-Cas9 mediated knock-ins, we generated knock-in cell lines where histone H2B (H2B) is labelled with mNeonGreen (mNG). HNEC3i basal cells were electroporated using two vectors, which contained mNeonGreen flanked by histone H2B-specific homology arms and H2B-specific guide RNA (gRNA) and Cas9 (figure 6a). 10 days after electroporation, clones with green fluorescent nuclei could be readily observed, indicating successful integration of mNeonGreen into the histone H2B gene (figure 6b, left panel). Although knock-in efficiency was low, ∼0.05–0.1%, we could enrich positive cells using FACS to generate a 100% HNEC3i H2B-mNG positive cell line (figure 6b, c). The knock-in cell line retained the ability to differentiate (figure 6d). Notably, we could observe two types of cells, one with low and one with high fluorescent levels, suggesting a mixed population of heterozygous and homozygous histone H2B-mNG knock-in cells, respectively (figure 6e). We conclude that knock-in airway epithelial cell lines can be generated using electroporation of two vectors which together contain the repair template, gRNA and Cas9, and that positive cells can be enriched by using FACS.

FIGURE 6.

CRISPR-Cas9 histone H2B-mNeonGreen knock-in in immortalised human primary nasal epithelial cells (HNEC3i). a) Schematic overview of CRISPR-Cas9 knock-in strategy. HNEC3i basal cells were electroporated with two vectors containing mNeonGreen (mNG) flanked by upstream (UHA) and downstream homology arms (DHA) and histone H2B-specific guide RNA (gRNA), Cas9 and puromycin (puro) resistance gene. Positive cells were enriched using fluorescence-activated cell sorting (FACS), expanded and differentiated in air–liquid interface (ALI) culture, and analysed using immunofluorescence staining. b) Representative EVOS microscope image of endogenous mNG signal in HNEC3i basal cells 10 days after electroporation and after FACS. c) Flow cytometry analysis of HNEC3i histone H2B-mNG. Cell selection based on mNG signal intensity (x-axis) and red autofluorescence was used to improve cell separation (y-axis). Cells with high mNG signal were collected. d, e) Immunofluorescence image of sorted and ALI differentiated HNEC3i H2B-mNG cells stained for α-tubulin (magenta), F-actin (white), 4′,6-diamidino-2-phenylindole (DAPI) (blue) and mNG (histone H2B-mNG) fluorescence (green), showing cross sections at the d) apical side or e) middle of the epithelium.

Discussion

In this study, primary human airway epithelial basal cells isolated from bronchial (HBEC) and nasal (HNEC) epithelium were successfully immortalised using lentiviral expression of hTERT to create the cell lines HBEC1i, HBEC2i, HNEC1i, HNEC2i and HNEC3i. The immortalised cell lines have been deposited in a biobank and are available to the research community on request. Immortalised cell lines retained the ability to differentiate into the major airway cell types, which displayed junctional integrity and ion transport function, and had motile cilia. Importantly, these airway characteristics remained stable in epithelial cells that were differentiated from immortalised basal cells of both early and late passages. Next, the validated, immortalised airway epithelial cell lines were used to optimise a genome editing toolbox for lentiviral overexpression of tagged proteins and CRISPR-Cas9 mediated genome editing.

Long-term airway epithelial cultures have been generated previously using hTERT expression in combination with the expression of oncoproteins or Cdk4, or co-culture with irradiated fibroblast feeder cells (reviewed in [13, 14]). In the current study, similar to previous work [20–22], hTERT alone was sufficient to immortalise basal cells. Basal cells in this study and in the work from the Crystal lab were grown in optimised cell culture medium, which by itself could aid to reduce cell stress and promote the proliferative capacity of basal cells [21, 22]. This idea is in line with recent work suggesting that low oxygen, presence of the Rho-associated coiled coil protein kinase (ROCK) inhibitor and irradiated fibroblast feeder cells can extend the proliferation limit of primary bronchial airway epithelial cells to >200 population doublings [12].

Immortalisation of human airway cells facilitates extended passaging and enables multiple rounds of selection, which are often essential for the validation of genome-edited cell lines. Here, we demonstrate with different genome editing workflows that a limited number of edited basal cells or a subset of labelled basal cells can be efficiently expanded to cryopreserve them, subsequently differentiate them and use them for functional assays. These techniques are instrumental to characterise the effects of complex, additive mutations, to conduct live-cell imaging studies or to visualise proteins in the absence of specific antibodies by knocking-in different tags. Furthermore, they offer means to model rare, disease-associated mutations in vitro, particularly when patient-derived material is scarce or unavailable, whereas in the past, such studies were mostly conducted in genetically engineered mouse models [39]. Moreover, donor heterogeneity in monogenic disease studies can be eliminated by employing isogenic lines or specifically addressed by inclusion of a panel of immortalised cell lines from multiple donors and origins. Lastly, these cell lines can provide ample material for large-scale assays for therapy development, including target identification and validation studies, and assessment of the effects of toxins or pharmacological agents on differentiated airway epithelial cells [40].

Taken together, we provided an easy-to-adopt genome editing toolbox which can help to address in vitro fundamental and disease-associated questions concerning airway epithelium without continuous dependency on donor material or animal models.

Supplementary material

Please note: supplementary material is not edited by the Editorial Office, and is uploaded as it has been supplied by the author.

Supplementary material

01708-2025.SUPPLEMENT