Primary Human Nasal Epithelial Cells: Biobanking in the Context of Precision Medicine

Summary

Here we describe the isolation, amplification, and differentiation of primary human nasal epithelial (HNE) cells at the air-liquid interface and a biobanking protocol allowing to successfully freeze and then thaw amplified HNE. The protocol analyzes electrophysiological properties of differentiated HNE cells and CFTR-related chloride secretion correction upon different modulator treatments.

Abstract

Human nasal epithelial (HNE) cells are easy to collect by simple, non-invasive nasal brushing. Patient-derived primary HNE cells can be amplified and differentiated into a pseudo-stratified epithelium in air-liquid interface conditions to quantify cyclic AMP-mediated Chloride (Cl^-) transport as an index of Cystic Fibrosis Transmembrane Conductance Regulator (CFTR) function. If critical steps such as quality of nasal brushing and cell density upon cryopreservation are performed efficiently, HNE cells can be successfully biobanked. Moreover, short-circuit current studies demonstrate that freeze-thawing does not significantly modify HNE cells' electrophysiological properties and response to CFTR modulators. In the culture conditions used in this study, when less than 2 x 10⁶ cells are frozen per cryovial, the failure rate is very high. We recommend freezing at least 3 x 10⁶ cells per cryovial. We show that dual therapies combining a CFTR corrector with a CFTR potentiator have a comparable correction efficacy for CFTR activity in F508del-homozygous HNE cells. Triple therapy VX-445 + VX-661 + VX-770 significantly increased correction of CFTR activity compared to dual therapy VX-809 + VX-770. The measure of CFTR activity in HNE cells is a promising pre-clinical biomarker useful to guide CFTR modulator therapy.

Introduction

Cystic Fibrosis (CF) is an autosomal recessive disorder resulting from mutations in the Cystic Fibrosis Transmembrane Conductance Regulator (CFTR) gene leading to the absence or dysfunction of the CFTR protein, an anion channel located at the apical surface of epithelia^1,2. Recent advances in CFTR therapy have improved the prognosis of the disease, and the last approved drugs combining CFTR correctors and CFTR potentiators led to major improvements in lung function and quality of life for CF patients carrying the most frequent mutation p.Phe508del mutation (F508del)^3,4. Despite this promising therapeutic progress, around 10% of CF patients are ineligible as they carry mutations that are unrescuable by these CFTR modulators. For these patients, there is a need to test other drugs or drug combinations to find the most efficient combination for specific mutations, highlighting the importance of personalized therapies.

Human nasal epithelial (HNE) cells are easy to collect by simple, non-invasive nasal brushing and allow quantification of cyclic AMP-mediated Chloride (Cl⁻) transport as an index of CFTR function. HNE cells yield an accurate model of human airway, but their lifespan is limited in culture. Thanks to the optimization of culture techniques, patient-derived primary HNE cells can be conditionally reprogrammed with Rho-associated kinase inhibitor (ROCKi), amplified, and differentiated into a pseudo-stratified epithelium in air-liquid interface (ALI) conditions on microporous filters^5,6. Numerous culture protocols for HNE culture exist (commercially available, serum-free, "homemade", co-culture with feeder-cells, etc.), and choice of media and culture conditions have been described to impact growth, cell population differentiation and epithelial function^7,8. The protocol here presents a simplified, feeder-free, ROCKi amplification method that allows to successfully obtain a large number of HNE cells that are then differentiated at ALI for CFTR function assays.

We have demonstrated that, in differentiated HNE cells, a 48 h treatment with CFTR modulators is sufficient to induce electrophysiological correction of CFTR dependent Cl^- current and that the correction observed in vitro may be correlated with the patient's clinical improvement⁹. HNE cells, therefore, represent an appropriate model not only for fundamental CF research but for pre-clinical studies with patient-specific CFTR modulator testing. In this context of personalized therapy, the goal of the protocol was to validate that cryopreserved HNE cells from CF patients, grown in our conditions, were an appropriate model for CFTR correction studies, and similar results could be expected when comparing CFTR dependent Cl^- transport from fresh and frozen-thawed cells. The study also assessed different CFTR modulators' efficacy when using dual and triple therapies.

Protocol

All experiments were performed following the guidelines and regulations described by the Declaration of Helsinki and the Huriet-Serusclat law on human research ethics.

1. Preparation of flasks and different media

1. Prepare amplification, air-liquid, and freezing medium as described in Table 1.
2. Prepare a stock solution of human collagen IV by dissolving 50 mg of collagen in 100 mL of 0.2% glacial acetic acid. Mix on a magnetic stirrer for at least 1 h, and filter sterilize the solution. Store at 4 °C for up to 6 months in a glass bottle, protected from light.
3. Prepare a working solution of collagen IV by diluting the stock solution 1:5 in double-distilled water. Store at 4 °C for up to 6 months.
4. Coat plastic cell culture flasks or transwell filters with the collagen working solution. Evenly distribute 100 µL for 0.33 cm² transwell filters, 500 µL for 25 cm² flasks, and 1 mL for 75 cm² flasks on the whole growth surface of the flask.
   NOTE: This can be achieved by gently inclining the flask from side to side. Alternatively, to facilitate coating the flasks, an increased volume can be used, and when the entire surface is evenly covered, collagen solution is aspirated to leave only the indicated volume. The aspirated collagen solution can then be used immediately to coat the next flask (i.e., for three 75 cm² flasks, distribute 3 mL of collagen solution evenly in the first flask, aspirate 2 mL, which are then used for the next flask, and so on).
5. Leave the cell culture flasks in the incubator for a minimum of 2 h or overnight. Aspirate the collagen thoroughly and leave the flasks to dry in the incubator or under the hood for a minimum of 20 min or overnight.
6. Wash with 10 mL of Mg²⁺- and Ca²⁺-free DPBS, allow to dry in the incubator, and store in aluminum foil to protect from light. Store the flasks for up to 6 months at room temperature (RT).

2. Nasal brushing

NOTE: Ensure that nasal brushing is performed while the participant does not have an acute infection. If CF patients present a pulmonary infection, refer to the patient's antibiogram and add additional antibiotics to the amplification medium. Human nasal epithelial cells were collected by nasal brushing from F508del/F508del patients as previously described⁹.

1. Ask the patient to blow their nose, then topically anesthetize the mucosa of both nostrils using cotton meshes soaked in xylocaine 5% with naphazoline solution.
2. Gently brush the medial wall and the inferior turbinate of both nostrils using a cytology brush. Soak the brush in a 15 mL tube with 2 mL of commercially available flushing medium; shake gently to detach cells. Repeat brushing in the second nostril.
3. Add the following antibiotics to the flushing medium: Penicillin (100 U/mL), Streptomycin (100 µg/mL), Tazocillin (10 µg/mL), Amphotericin B (2.5 µg/mL) and Colimycin (16 µg/mL).
   ​NOTE: Nasal brushings can be shipped at RT to dedicated labs for expansion and culture and can be seeded up to 72 h after brushing.

3. Isolation of HNE

NOTE: HNE were isolated from the cytology brush, washed with Mg²⁺- and Ca²⁺-free DPBS, dissociated with 0.25% Trypsin, and seeded onto a 25 cm² plastic flask as previously described¹⁰.

1. Detach cells from the cytology brush by repeatedly passing the brush through a 1000 µL pipette cone and rinsing with 2 mL of Mg²⁺- and Ca²⁺-free DPBS. Centrifuge at 500 x g for 5 min at 4 °C and discard the supernatant.
2. Resuspend the pellet in 0.25% Trypsin for 8-12 min to dissociate cells. Stop the enzymatic reaction by adding 5 mL of the amplification medium.
3. Centrifuge at 500 x g for 5 min at 4 °C and discard the supernatant. Resuspend the pellet in 8 mL of the amplification medium and seed onto a 25 cm² collagen-coated flask. Grow at 37 °C and 5% CO₂. This corresponds to passage 0.
4. The following day, replace the medium with 5 mL of fresh amplification medium and monitor the cells daily for attachment, morphology (cells should have a cohesive cobblestone appearance), and number.
   ​NOTE: The initial seeding density varies depending on the quality of the nasal brushing and the proportion of epithelial cells. Freshly isolated HNE cells usually reach 80%-90% confluency within 3-10 days. A slower growth rate is suggestive of insufficient epithelial cells in the sample and should be discarded.

4. Amplification and passaging of human nasal epithelial cells

1. Grow human nasal epithelial cells in the amplification medium at 37 °C and 5% CO₂ on collagen-coated flasks until 80%-90% confluency, changing medium every 48-72 h. For 25 cm² flasks, use 5 mL of amplification medium and 10 mL for 75 cm² flasks.
2. When cells reach 80%-90% confluency, wash the cells with 10 mL of Mg²⁺- and Ca²⁺-free DPBS, aspirate, and discard.
3. Add 1 mL of 0.25% Trypsin for a 25 cm² flask (or 2 mL of Trypsin for a 75 cm² flask) and return to the incubator (37 °C, 5% CO₂) for 8-12 min.
4. Tap the flask firmly, with the palm of the hand, to help the cells detach.
5. Add 10 mL of the amplification medium to stop the enzymatic reaction. Vigorously rinse the flask, using the 10 mL pipette to draw up and expel the amplification medium over the flask surface to rinse and detach cells.
6. Centrifuge at 500 x g for 5 min at 4 °C and discard the supernatant.
7. Resuspend the pellet in 5-10 mL of the amplification medium.
8. Count the cells using a hemocytometer.
   ​NOTE: Cells can be frozen at this point (see steps 5.1-5.3). If further amplification is required, re-seed onto collagen-coated flasks (a 25 cm² flask can be expanded into three 75 cm² flasks, a greater dilution is not recommended).

5. Cryopreservation of primary nasal epithelial cells

NOTE: Grow HNE cells until passage 1 before biobanking to obtain enough cells to facilitate cell growth when thawed. Biobanking is, however, possible at initial passage 0 or passage 2. The following cryopreservation steps are adapted to all passages.

1. After cell counting, centrifuge at 500 x g for 5 min at 4 °C, and discard the supernatant
2. Resuspend the pellet in the enriched freezing medium (Table 1) to obtain 3 x 10⁶-5 x 10⁶ cells per mL and per cryovial.
3. Slowly freeze the cells by reducing the temperature at ~1 °C per min in an appropriate cryo-freezing container at -80 °C. The next day, move the preserved sample to a nitrogen storage container for long-term storage (the present study is limited to 17 months storage).

6. Thawing frozen amplified HNE cells

1. Warm up the water bath to 37 °C. Prepare the amplification medium as described in Table 1 and warm the medium to 37 °C.
2. Remove cryovials from the nitrogen storage tank and rapidly place them in the water bath, taking care not to submerge the whole vial in the water. Remove the cryovials from the water bath when only a small frozen droplet remains. Wipe the vial with 70% alcohol, wipe dry, and place under the hood.
3. Using a 1 mL pipette, transfer the thawed cells to an empty 15 mL tube.
4. In a drop-wise manner, add 1 mL of warm amplification medium. After 1 min, add another 1 mL of amplification medium and wait an additional minute.
5. Add 10 mL of the amplification medium and centrifuge for 2 min at 500 x g.
6. Aspirate and discard the supernatant. Resuspend the cell pellet in a volume of amplification medium required to achieve a cell density of at least 1 x 10⁶ cells/mL.
7. Seed the cells onto a collagen-coated flask containing the amplification medium.
   1. If the cryovial contains more than 4 x 10⁶ cells, seed onto a 75 cm² flask, in 10 mL of the amplification medium
   2. If cryovial contains less than 4 x 10⁶ cells, seed cells onto a 25 cm² flask, in 5 mL of the amplification medium
8. Incubate at 37 °C, 5% CO_2, and visually observe cell expansion over the next 2-3 days.
9. Then, perform cell amplification as detailed in section 4.
   ​NOTE: If few cells were harvested in step 4.7. (i.e., if freezing at passage 0 before expansion), steps 6.5. and 6.6. can be omitted, and cells can be seeded directly onto a 25 cm² collagen-coated flask without centrifugation. The medium must then be changed the following day to remove dimethyl sulfoxide (DMSO).

7. Differentiation of human nasal epithelial cells at the air-liquid interface

NOTE: HNE were differentiated at the air-liquid interface as previously described¹⁰.

1. Seed the cells at a density of 330,000 cells/filter on 0.33 cm² collagen-coated porous filters and supplement with 300 µL of the amplification medium at the apical side and 900 µL of the air-liquid medium at the basolateral side.
2. After 3 days, aspirate the apical medium and culture the cells at the air-liquid interface for 3-4 weeks in the air-liquid medium to establish a differentiated epithelium. Change the basal medium every 48-72 h.

8. CFTR modulators

1. Prepare air-liquid medium containing CFTR modulators. Use VX-445, VX-661, and VX-809 at a final concentration of 3 µM and VX-770 at 100 nM. Use corrector ABBV-2222 at 1 µM final concentration and potentiator ABBV-974 at 10 µM. Use 100% DMSO to dissolve all CFTR modulators.
2. Prepare air-liquid medium containing the same amount of 100% DMSO as in corrector medium.
3. Add the medium containing drugs or DMSO to the basolateral side of polarized HNE cells grown at an air-liquid interface and incubate for 24 h in 5% CO₂ at 37 °C.
4. After 24 h, aspirate and discard the medium and replace with fresh medium prepared as in steps 8.1-8.2., and further incubate for a 24 h period in 5% CO₂ at 37 °C.

9. Ussing chamber studies

1. Measure short-circuit current measurements (Isc) under voltage-clamp conditions as previously described⁹.
   NOTE: Transepithelial electrical resistance (TEER) of cultures was measured with a chopstick voltmeter, and only cultures reaching at least 200 Ω·cm² were considered for the following experiments. Filters of polarized HNE cells were mounted in Ussing chambers, and short-circuit current measurements (Isc) were measured under voltage-clamp conditions.

10. Immunocytochemistry

1. Perform Immuno-detection as previously described⁹.
   ​NOTE: CFTR immuno-detection was performed using the anti-CFTR C-terminal (24-1) monoclonal antibody overnight at a 1/100 dilution. Zona-occludens-1 (ZO-1) (1/500 dilution), alpha-tubulin (1/300 dilution), Muc 5AC (1/250 dilution) and cytokeratin 8 (1/250 dilution) staining were done on additional filters. After washing with PBS-Triton-X100 0.1%, goat secondary antibodies conjugated to Alexa 488 or 594 were added for 30 min in 10% goat serum (1/1000 dilution). After a final wash, filters were cut from support and mounted with Vectashield mounting medium containing DAPI. A confocal microscope (63x/1.4 oil differential interference contrast λ blue PL APO objective) was used to capture images, which were analyzed with the ImageJ software.

11. Statistical analysis

1. Perform statistical analysis using appropriate software.
   NOTE: Statistical analysis was performed using S.A.S software. As several HNE filters were obtained per patient and per condition, quantitative parameters were expressed as median values (± SEM) per patient. Comparisons (mean ± SD) were carried out using Wilcoxon matched pairs signed rank test or unpaired t-test.

Representative Results

Fresh HNE cells cultured at the air-liquid interface display typical features of the polarized and differentiated respiratory epithelium as assessed by immunostaining (Figure 1). HNE cells re-differentiate into a heterogeneous layer of epithelial cells (positive keratin 8 immunostaining) that mimic the in vivo situation of a pseudo-stratified respiratory epithelium composed of ciliated (positive alpha-tubulin staining) and non-ciliated mucus-producing goblet cells (positive Muc5Ac immunostaining). The overall epithelium presents tight junctions (assessed by ZO-1, zona-occludens-1 staining). Strong apical CFTR staining is visible in wild-type (WT) HNE cells, and decreased CFTR staining both at the apical surface and in the cytoplasm is observed in F508del/F508del HNE.

We considered positive outcomes as samples not only successfully expanding but also successfully differentiating into a pseudo-stratified epithelium after 3 weeks at air-liquid interface and presenting a TEER higher than 200 Ω·cm², allowing short-circuit current measurements. As only a few dedicated labs perform CFTR-related Cl^- secretion assays by short-circuit current measurements, samples often require transport from hospitals to dedicated facilities.

In the 78 fresh HNE cell samples, success rates were compared in nasal brushings seeded immediately and in brushings seeded after 1-7 days of transport in flushing medium at room temperature. The majority of samples (55%) were seeded 1 day after brushing, and 82% of samples were seeded within 48 h. The mean percentage of successful cultures in all 78 samples was 82% and reached 87% in samples seeded between day 0 and day 5 (Table 2).

The success and failure rates for identical samples differentiated in both fresh and freeze-thawed conditions were compared. 83% of samples that differentiated successfully in fresh conditions also were successful in cryopreserved samples. Likewise, in samples that failed to differentiate in fresh conditions, 71% were also unsuccessful in freeze-thaw conditions (Table 2). All these data suggest that the quality of nasal brushing is the key factor to a successful culture.

We also wanted to determine the optimal number of cells per cryovial. Samples frozen at passage 0, from a single 25 cm² generally only allow to reach 1-1.5 x 10⁶ cells per flask. Samples frozen at passage 1 and amplified to three 75 cm² flasks generally allowed to reach 10 x 10⁶ cells, thus allowing to freeze several vials containing at least 3 x 10⁶ cells. In the culture conditions described in this protocol, 50% of samples frozen between 2 x 10⁶ and 3 x 10⁶ cells per cryovial failed to grow when thawed, reaching 80% when below 2 x 10⁶ cells per cryovial (Table 2).

CFTR-related Cl^- secretion reflects CFTR function and can be measured by short circuit current (Isc) experiments in Ussing chambers. The Isc variation in response to Forskolin (ΔIsc_Forskolin) and CFTR inhibitor inh172 (ΔIsc_CFTRinh172) are the main endpoints for CFTR corrector efficacy assessment.

Short-circuit-current experiments were performed on either fresh HNE, expanded in amplification medium and seeded onto filters at passage 2 and grown at air-liquid interface for 3 weeks in air-liquid medium, or from HNE that had been previously frozen at passage 1 in enriched freezing medium, thawed, expanded, and seeded on porous filters at passage 3. Cryopreservation times ranged from 3 months to 16 months, with a mean of 9.8 months. In the culture and cryopreservation conditions detailed in this study, no significant changes in growth or morphology were observed between different passages or fresh or frozen-thawed F508del/F508del HNE cultures.

For both mean ΔIsc_Forskolin (Figure 2A) and mean ΔIsc_CFTRinh172 (Figure 2B) in DMSO treated HNE cells, no significant difference was observed between fresh or frozen-thawed HNE. To assess whether functional correction of CFTR was influenced by cryopreservation, mean Isc changes were measured in VX-809 + VX-770 treated cells compared to DMSO treated HNE cells in fresh (HNE cells derived from n =78 patients) or frozen-thawed HNE (HNE cells derived from n = 9 patients) (Figure 2). Both frozen-thawed and fresh cells displayed significant correction upon treatment, with a significant increase in ΔIsc_Forskolin (p < 0.05) (Figure 2A), and a significant decrease in the ΔIsc_CFTRinh172 (p < 0.05) (Figure 2B). The mean fold increase for ΔIsc_Forskolin (ΔIsc_Forskolin VX-809 + VX-770/ΔIsc_Forskolin DMSO) reached 2.9 ± 1.6 in frozen-thawed HNE cells and was not significantly different from the fold increase in fresh HNE cells (2.5 ± 2.0). Similarly, the fold decrease in ΔIsc_CFTRinh172 (ΔIsc_CFTRinh172 VX-809 + VX-770/ΔIsc_CFTRinh172 DMSO) was not significantly affected by cryopreservation: 1.49 ± 0.9 in frozen-thawed HNE compared to 2.5 ± 3.7 in fresh HNE.

The functional correction activity of several modulator therapies was further investigated to assess whether this HNE cell model could discriminate between different treatments. The results here combine HNE cells from both fresh and frozen-thawed conditions as the response to treatment was not significantly altered by cryopreservation (Figure 2).

We first assessed the corrector response of two approved CFTR modulator therapies on Cl^- transport after a 48 h treatment with either VX-809 (3 µM) + VX-770 (100 nM) or VX-661 (3 µM) + VX-770 (100 nM) in HNE cells from 10 F508del/F508del patients (Figure 3). As compared to DMSO treated HNE cells, ΔIsc_Forskolin (Figure 3A) and ΔIsc_CFTRinh172 (Figure 3B) were significantly improved by both VX-809 + VX-770 and VX-661 + VX-770. Both treatments functionally corrected CFTR-related Cl^- secretion to a similar level, with a significant mean fold increase in ΔIsc_Forskolin of 1.75 ± 1.39 for VX-809 + VX-770 (p < 0.001) and 0.97 ± 0.93 for VX-661 + VX-770 (p < 0.01). Mean ΔIsc_CFTRinh172 fold decrease was 1.79 ± 2.04 for VX-809 + VX-770 (p < 0.001) and was not significantly different from the mean ΔIsc_CFTRinh172 fold decrease of 1.16 ± 1.44 for VX-661 + VX-770 (p < 0.001 vs. DMSO).

Three different CFTR modulator therapies in HNE from six F508del/F508del patients (Figure 4) were further compared. The results here combine HNE cells from both fresh and frozen-thawed conditions. Both Vertex and Abbvie CFTR modulators, when used as a combination of a CFTR corrector and a CFTR potentiator, corrected ΔIsc_Forskolin (Figure 4A) and ΔIsc_CFTRinh172 (Figure 4B) to a similar extent. The partial correction observed in Figure 3 by VX-809 + VX-770 was confirmed in these HNE cells with a mean ΔIsc_Forskolin fold increase of 2.87 ± 1.67 (p < 0.05). ABBV-2222 (1 µM) + ABBV-974 (10 µM) treatment induced a similar correction representing 91% of VX-809 + VX-770 correction for ΔIsc_Forskolin (mean ΔIsc_Forskolin fold increase of 2.6 ± 1.4, p < 0.05 vs. DMSO) and 85% of VX-809 + VX-770 for ΔIsc_CFTRinh172 (mean ΔIsc_CFTRinh172 fold decrease of 4.2 ± 5.7, p < 0.05 vs. DMSO)_.

Interestingly, when HNE were treated with the triple combination VX-445 (3 µM) + VX-661 (3 µM) + VX-770 (100 nM), correction efficacy significantly improved, reaching 312% of VX-809 + VX-770 ΔIsc_Forskolin correction (mean ΔIsc_Forskolin fold increase of 9.0 ± 4.1, p < 0.05 vs. VX-809 + VX-770) and 372% of ΔIsc_CFTRinh172 (mean ΔIsc_CFTRinh172 fold decrease of 15.5 ± 10.5, p < 0.05 vs. VX-809 + VX-770).

Figure 1: Differentiation markers in HNE cultures. Representative confocal microscopy images of immune-fluorescent staining of CFTR (green), Zona-Occludens-1 (ZO-1, red), alpha-tubulin (green), Muc5AC (red) and cytokeratin 8 (K8, green) in wild-type (WT) (first panel) and F508del (panels 2 -4) HNE cells grown at air-liquid interface for 3 weeks. Nuclei were stained blue with DAPI. Projections of confocal microscopy images, top view (upper panels) and side view (lower panels) of three-dimensional (3D) Z stacks reconstitutions. Scale bar (20 µm) applies to both upper and lower panels. Please click here to view a larger version of this figure.

Figure 2: Effect of freezing-thawing on CFTR-related Cl^- secretion in HNE cells. Summary of short-circuit current (Isc) recordings (mean ± SD) in HNE cultures derived from F508del/F508del patients and grown at air-liquid interface for 3 weeks. Cells were treated for 48 h with vehicle alone (DMSO) or VX-809 (3 µM) + VX-770 (100 nM) in fresh (light grey, patient-derived HNE cells from 78 patients) or frozen-thawed (dark grey, patient-derived HNE cells from 9 patients). Data represent the mean (± SD) response to acutely added (A) Forskolin (10 µM)/3-isobutyl-1-methylxanthine (IBMX, 100 µM), ΔIsc_Forskolin or (B) CFTR inhibitor inh172 (5 µM) ΔIsc_CFTRinh172. A minimum of two filters were analyzed per patient and per condition. Asterisks indicate the statistical difference between correctors and control (DMSO-treated) * p < 0.05 (unpaired t-test). Please click here to view a larger version of this figure.

Figure 3: Correction of CFTR function by dual therapies. Summary of short-circuit current (Isc) recordings (mean ± SD) in HNE cultures derived from 10 F508del/F508del patients and grown at air-liquid interface for 3 weeks (combined results from both fresh and frozen-thawed cultures). Cells were treated for 48 h with vehicle alone (DMSO); VX-809 (3 µM) + VX-770 (100 nM) or VX-661 (3µM) + VX-770 (100 nM). Data represent the mean (± SD) response to acutely added (A) Forskolin (10 µM)/IBMX (100 µM), ΔIsc_Forskolin or (B) CFTR inhibitor inh172 (5 µM), ΔIsc_CFTRinh172. A minimum of two filters were analyzed per patient and per condition. Asterisks indicate a statistical difference between correctors and control (DMSO-treated) ** p < 0.01, *** p < 0.001, **** p < 0.0001 (Wilcoxon matched pairs signed rank test). Please click here to view a larger version of this figure.

Figure 4: Efficiency of triple-therapy vs. dual therapies on CFTR correction. Summary of short-circuit current (Isc) recordings (mean ± SD) in HNE cultures derived from 6 F508del/F508del patients and grown at air-liquid interface for 3 weeks (combined results from both fresh and frozen-thawed cells). Cells were treated for 48 h with vehicle alone (DMSO); VX-809 (3 µM) + VX-770 (100 nM); ABBV-2222 (1 µM) + ABBV-974 (10 µM); or VX-661 (3 µM) + VX-445 (3 µM) + VX-770 (100 nM). Data represent the mean (± SD) response to acutely added (A) Forskolin (10 µM)/IBMX (100 µM), ΔIsc_Forskolin or (B) CFTR inhibitor inh172 (5 µM), ΔIsc_CFTRinh172. A minimum of two filters were analyzed per patient and per condition. Asterisks indicate a statistical difference between treatments. * p < 0.05 (Wilcoxon matched pairs signed rank test). Please click here to view a larger version of this figure.

Media Component	Final concentration
AMPLIFICATION MEDIUM
Advanced DMEM/F-12 Nutrient mixture	86%
Fetal Bovine Serum	10%
Penicillin	100 U/mL
Streptomycin	100 µg/mL
Tazocillin	10 µg/mL
Amphotericin B	2.5 µg/mL
Colimycin	16 µg/mL
Ciprofloxacin	3 µg/mL
Hydrocortisone	0.2 µM
Insulin	5 µg/mL
Epinephrine	0.5 µg/mL
EGF	10 ng/mL
Y-27632 (Rho-associated kinase inhibitor)	10 µM
AIR-LIQUID MEDIUM
Advanced DMEM/F-12 Nutrient mixture	94%
UltroserG	2%
Penicillin	100 U/mL
Streptomycin	100 µg/mL
Tazocillin	10 µg/mL
Amphotericin B	2.5 µg/mL
Colimycin	16 µg/mL
ENRICHED FREEZING MEDIUM
F12 Nutrient Mixture	77%
HEPES	3%
Fetal Bovine Serum	10%
DMSO	10%

Table 1: Amplification, air-liquid, and enriched freezing medium composition. List of reagents and corresponding final concentrations.

Length of conservation of nasal brushing before seeding (n = 78)
Days before seeding	Success rates (%)
0	83
1	84
2	67
3	100
4	89
5	100
7	0
Outcomes from the same initial nasal brushing grown in both fresh and frozen-thawed conditions (n = 13)
Outcome in fresh conditions	Success rates upon freeze-thawing (%)
success	83
failure	29
Impact of cell number on success rates (n = 36)
Number of cells per cryovial	Success rates upon freeze-thawing (%)
<2 x 106	20
2–3 x 106	50
3–4 x 106	67
>4 x 106	79

Table 2: Parameters influencing success rates in fresh and frozen-thawed HNE cell cultures. Success rates were defined as nasal brushings that successfully expanded and differentiated into a pseudo-stratified epithelium after 3 weeks at the air-liquid interface and presented a TEER higher than 200 Ω·cm². Before seeding, samples were shipped at room temperature in flushing medium with antibiotics. Results represent data from 78 fresh samples and 36 frozen-thawed samples. 13 nasal brushings were analyzed in both fresh and frozen-thawed conditions.

Discussion

The use of patient-derived nasal epithelial cells as surrogates for human bronchial epithelial (HBE) cells to measure CFTR activity in the context of personalized medicine has been proposed as HNE reproduce cells' properties in culture^9,11. The strong advantage of HNE over HBE cell cultures is that they are easily and non-invasively sampled. Short-circuit current measurements in HNE cell cultures enable assessment of CFTR-dependent Cl^- transport across the epithelium, which was shown to significantly correlate to CFTR in vivo activity assessed by nasal potential difference in patients with various genotypes⁹. The rescue of CFTR activity in HNE cells upon VX-809 treatment also significantly correlated to the improvement of clinical outcome FEV₁ in lumacaftor-ivacaftor treated F508del homozygous patients¹².

Importantly, this study shows that freshly harvested HNE cells easily tolerate transport and standard shipping conditions, (i.e., they survive 72 h at room temperature) which allows the testing of patient's cells originating from various geographical locations. The critical factors for successful culture are mainly the quality of nasal brushing and the number of living cells at sampling. A number of protocols have been published on growing and differentiating HNE cell cultures; some with no conditional reprogramming during the expansion phase^{13,14,15,16,17,18,19,20}, some with conditional reprogramming in feeder^{9,11,12,21,22,23,24,25,26} or feeder-free^5,8,27,28,29 conditions, and several comparing different methods^7,8. They all allow obtaining a relatively large number of ALI cultures with correct differentiation and polarization as well as good TEER values. Similar to this study, many used cryopreserved HNE cells frozen at passage 1^13,28. Interestingly, several authors report that ROCKi can improve survival after cryopreservation when added to the post-thaw culture medium and increase the number of cells that remain attached^5,29, suggesting that the amplification medium used in this study is particularly appropriate for cryopreservation.

Here, the study shows that HNE cells can be successfully biobanked, the critical factor again being the quality of the nasal brushing. A good correlation between the outcome of fresh samples and frozen-thawed ones is shown, suggesting that in order to optimize the quality of HNE cells that are going to be biobanked for further studies, a preliminary study could be performed on fresh samples if they fail to achieve correct ALI differentiation and acceptable TEER value; a second nasal brushing would be preferable rather than freezing cells that have a high probability of poor outcome. A second critical factor for cryopreservation is cell number; in the culture conditions used in this study, when less than 2 x 10⁶ cells are frozen per cryovial, the failure rate is very high. Freezing at least 3 x 10⁶ cells per cryovial is recommended.

The results in this study suggest that freeze-thawing does not significantly modify HNE cells' electrophysiological properties or response to CFTR modulators. The range of measures in HNE cell cultures from the same F508del-homozygous patient obtained without freezing or after thawing was not different upon VX-809 + VX-770 treatment. Similar results were previously reported for HBE cells where Forskolin-stimulated Isc were stable after several thaw/culture cycles of non-treated F508del cells³⁰. This provides evidence that HNE cells can be frozen and stored in a biobank and studied later.

An increasing number of studies show that HNE cell cultures can be used to test the efficacy of different therapies in the context of personalized medicine. The analysis showed that dual therapies combining a corrector with a potentiator all have comparable correction efficacy for CFTR-dependent Cl^- secretion in F508del-homozygous HNE cells. However, an important interpatient variability was noticed, confirming published results^9,12,20,25. By contrast, VX-445 + VX-661 + VX-770 tripled correction of CFTR activity compared to VX-809 + VX-770. A similar level of correction of F508del-CFTR in HNE cells upon triple combination was also observed by Laselva et al.²⁸. As reported by Veit et al.²⁶, correction of Cl^- channel function upon VX-445 + VX-661 + VX-770 treatment was even higher, reaching 62% of mean WT-CFTR activity (range of 19-36 µA/cm² in corrected F508del-HNE and 14-50 µA/cm² in WT-HNE)²⁶.

HNE cell cultures also served to test the efficacy of CFTR modulators for other class II mutations. An important improvement of CFTR activity upon VX-445 + VX-661 + VX-770 was observed in HNE ALI cultures harboring G85E/G85E, V520F/1717-1G>A, Y569/Y569D, M1101K/M1101K, and N1303K/N1303K genotypes²⁶. Similarly, Laselva et al.²⁸ showed a significant correction of CFTR with M1101K and N1303K mutations but no significant correction in G85E HNE cells. Galapagos/AbbVie correctors were evaluated for these lumacaftor-ivacaftor-resistant mutants³¹. AbbVie corrector AC2-1 and especially bi-corrector AC1+AC2-1 effectively improved Cl^- secretion in M1101K and G85E HNE cells, whereas N1303K derived cells were corrected by a combination of AC1+AC2-2.

Measures of CFTR activity in HNE cells are promising pre-clinical biomarkers predictive for the patient's response to tested treatment. Evidence that freezing-thawing cycles do not alter the correction level provides evidence that these cells can be biobanked and used as a powerful tool in the context of personalized medicine programs. A unique nasal brushing could therefore allow for sufficient material to compare drug efficacies, and patient's HNE cells could be thawed as new drugs become available with no need to resample the patients; this is particularly interesting in infants.

Materials

Name	Company	Catalog Number	Comments
ABBV-2222	Selleckchem	S8535
ABBV-974	Selleckchem	S8698
Advanced DMEM/F-12	Life Technologies	12634010
Alexa 488 goat secondary antibody	Invitrogen	A11001
Alexa 594 goat secondary antibody	Invitrogen	A11012
Amphotericin B	Life Technologies	15290026
Anti-alpha-tubulin antibody	Abcam	ab80779
Anti-CFTR monoclonal antibody (24-1)	R&D Systems	MAB25031
Anti-cytokeratin 8 antibody	Progen	61038
Anti-Muc5AC antibody	Santa Cruz Biotech	sc-20118
Anti-ZO-1 antibody	Santa Cruz Biotech	sc-10804
Ciprofloxacin		provided by Necker Hospital Pharmacy
Colimycin	Sanofi	provided by Necker Hospital Pharmacy
Collagen type IV	Sigma-Aldrich Merck	C-7521
cytology brush	Laboratory GYNEAS	02.104
DMSO	Sigma-Aldrich Merck	D2650
EGF	Life Technologies	PHG0311
Epinephrin	Sigma-Aldrich Merck	E4375
F12-Nutrient Mixture	Life Technologies	11765054
FBS	Life Technologies	10270106
Ferticult	Fertipro NV	FLUSH020
Flasks 25	Thermo Scientific	156.367
Flasks 75	Thermo Scientific	156.499
Glacial acetic acid	VWR	20104.298
HEPES	Sigma-Aldrich Merck	H3375
Hydrocortisone	Sigma-Aldrich Merck	SLCJ0893
Insulin	Sigma-Aldrich Merck	I0516
Mg2+ and Ca2+-free DPBS	Life Technologies	14190094
Penicillin/Streptomycin	Life Technologies	15140130
Tazocillin	Mylan	provided by Necker Hospital Pharmacy
Transwell Filters	Sigma-Aldrich Merck	CLS3470-48EA
Triton-X100	Sigma-Aldrich Merck	T8787
Trypsin 0,25%	Life Technologies	25200056
Vectashield mounting medium with DAPI	Vector Laboratories	H-1200
VX-445	Selleckchem	S8851
VX-661	Selleckchem	S7059
VX-770	Selleckchem	S1144
VX-809	Selleckchem	S1565
Xylocaine naphazoline 5%	Aspen France	provided by Necker Hospital Pharmacy
Y-27632	Selleckchem	S1049