In utero and postnatal ivacaftor/lumacaftor therapy rescues multiorgan disease in CFTR-F508del ferrets

Abstract

Cystic fibrosis (CF) is caused by mutations in the CF transmembrane conductance regulator (CFTR) gene, with F508del being the most prevalent mutation. The combination of CFTR modulators (potentiator and correctors) has provided benefit to CF patients carrying the F508del mutation; however, the safety and effectiveness of in utero combination modulator therapy remains unclear. We created a F508del ferret model to test whether ivacaftor/lumacaftor (VX-770/VX-809) therapy can rescue in utero and postnatal pathologies associated with CF. Using primary intestinal organoids and air-liquid interface cultures of airway epithelia, we demonstrate that the F508del mutation in ferret CFTR results in a severe folding and trafficking defect, which can be partially restored by treatment with CFTR modulators. In utero treatment of pregnant jills with ivacaftor/lumacaftor prevented meconium ileus at birth in F508del kits and sustained postnatal treatment of CF offspring improved survival and partially protected from pancreatic insufficiency. Withdrawal of ivacaftor/lumacaftor treatment from juvenile CF ferrets reestablished pancreatic and lung diseases, with altered pulmonary mechanics. These findings suggest that in utero intervention with a combination of CFTR modulators may provide therapeutic benefits to individuals with F508del. This CFTR-F508del ferret model may be useful for testing therapies using clinically translatable endpoints.

Keywords: Bacterial infections; Genetic diseases; Pulmonology; Therapeutics.

Figure 1. Generation of F508del-knockin ferret and characterization of CFTR mRNA expression.

(A) Schematic of the targeting strategy at the ferret CFTR exon 11 locus using Cas9/gRNA complex and oligonucleotide mutagenesis template targeting deletion of F508. Forward (F) and reverse (R) primers used for amplification of the locus are marked by arrows. (B) Sequencing results of PCR-amplified genomic DNA, using F/R primers shown in A, from a heterozygous F508del founder. The sequence confirms deletion of the nucleotide triplet (CTT), which resulted in F508del. (C) Two-step genotyping protocol used to differentiate WT, CFTR-F508del heterozygous (Het), and CFTR-F508del homozygous (Homo) offspring. Step 1: The first Surveyor step distinguished a heterozygous genotype from homozygous and WT. Only the heterozygous PCR product is cleaved during the Surveyor assay due to the mismatch in bases (marked with X), giving rise to 2 bands on agarose gels. Step 2: To distinguish homozygous from WT genotypes, reference WT and test sample genomic DNA are mixed at a ratio of 0.5:1 (WT/Test) and PCR is then performed with F/R primers. These PCR products are then used in a second Surveyor assay. Only the homozygous PCR product is cleaved during the Surveyor assay due to the mismatch in bases (marked with X), giving rise to 2 bands on agarose gels. (D) qPCR analysis for CFTR mRNA on ferret intestinal organoids generated from CFTR-KO (n = 4 donors), CFTR-WT (n = 8 donors), and CFTR-F508del homozygous (n = 3 donors) ferrets. Data represent mean ± range. (E) qPCR analysis for CFTR mRNA from proximal intestine (n = 5 donors), trachea (n = 4 donors for WT and n = 5 donors for F508del), and lung (n = 5 donors) from CFTR-WT and CFTR-F508del newborn ferrets. Data represent mean ± SEM.

Figure 2. Ferret F508del CFTR processing in intestinal organoids and differentiated air-liquid interface (ALI) cultures of airway epithelia.

(A) Western blots of CFTR and β-tubulin protein expression from ferret intestinal organoids of WT, CFTR-KO (KO), and CFTR-F508del homozygous genotypes. Band C indicates the mature (complex glycosylated) form of CFTR and Band B indicates the immature (core glycosylated) form of CFTR. (B) Western blots of CFTR and β-tubulin protein expression from differentiated airway epithelia grown at an ALI generated from WT and F508del (homozygous) ferret tracheal basal cells. F508del cells were treated overnight with either vehicle (DMSO) or LUM (3 μM). (C) Western blot of WT and F508del ferret intestinal organoid lysates incubated with or without endoglycosidase H (EndoH) to cleave N-linked glycosylated residues added in the ER. Susceptibility to EndoH cleavage indicates ER-resident CFTR (Band B) and produces the slightly smaller Band A form of CFTR. The fully mature form of CFTR (Band C) is not sensitive to EndoH. The majority of F508del CFTR protein is thus resident in the ER. (D) F508del ferret organoids were incubated at 37°C for 16 hours in the presence of vehicle (DMSO), LUM (CFTR corrector), IVA (CFTR potentiator) and LUM, or ELX and TEZ (CFTR correctors) at 3 μM final concentration. (E) F508del homozygous differentiated airway cultures grown at an ALI were incubated with LUM (3 μM, 16 hours at 37°C) and then evaluated by Western blotting. Three independent donor animals were used per condition. (F–H) Quantification of the relative intensity of CFTR forms normalized to β-tubulin in differentiated airway cultures treated with vehicle (DMSO) or LUM: (F) (Band C + Band B)/β-tubulin, (G) Band C/β-tubulin, and (H) Band B/β-tubulin. Data represent mean ± SEM. Statistical significance evaluated by paired 2-tailed Student’s t test (n = 6 Transwells, reflecting 2 replicates from each of 3 independent donors). P values are given in each panel.

Figure 3. Combination LUM/IVA treatment enhances forskolin-induced swelling responses of ferret CFTR-F508del intestinal organoids.

(A) Forskolin-induced swelling (FIS) assays on homozygous CFTR-F508del newborn ferret intestinal organoids in the presence of DMSO, IVA (3 μM), LUM (3 μM), or LUM/IVA (3 μM each). Each point represents the average area under the curve (AUC) for n = 17 cultures, reflecting 1–3 replicates from 14 donors. (B) FIS assays on homozygous CFTR-F508del adult ferret intestinal organoids in the presence of DMSO, IVA (3 μM), LUM (3 μM), or LUM/IVA (3 μM each) (n = 5 cultures, reflecting 1–2 replicates from 3 donors). (C and D) FIS assay on homozygous CFTR-F508del newborn ferret intestinal organoids in the presence and absence of (C) GlyH101 (n = 4 cultures, reflecting 1–2 replicates from 3 donors) and (D) DIDS (n = 5 cultures, reflecting 1–2 replicates from 4 donors). All data represent mean ± SEM. Significant differences were determined by 2-way ANOVA with Bonferroni’s post hoc test. *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001 for the following comparisons: (A and B) IVA, LUM, and LUM/IVA treatment groups versus DMSO controls; (C) LUM/IVA versus LUM/IVA/GlyH101 or DMSO versus GlyH101 (no significant differences); and (D) LUM/IVA versus LUM/IVA/DIDS or DMSO versus DIDS, which demonstrated no significant differences.

Figure 4. Short circuit (Isc) measurements of ferret CFTR-F508del differentiated airway epithelia grown at an air-liquid interface.

Differentiated airway epithelia were generated using tracheal basal cell cultures derived from CFTR-KO (KO), homozygous CFTR-F508del (F508del), and WT ferrets and polarized at an air-liquid interface (ALI) for 3 weeks. Ussing chamber measurements of Isc were performed under (A) bicarbonate-free symmetric-chloride buffer to assess chloride currents and (B) chloride-free symmetric-bicarbonate buffer to assess bicarbonate current. Graphs show the change in Isc (ΔIsc) following the sequential addition of amiloride, 4,4′-diisothiocyanatostilbene-2,2′-disulfonic acid (DIDS), forskolin and 3-isobutyl-1-methylxanthine (F&I), and GlyH101 (CFTR inhibitor). Data represent the mean ± SEM for the indicated number (n) of Transwells measured for each condition (6–8 Transwells assayed for each of 2–3 donors per genotype). Significant differences were determined by 1-way ANOVA with Bonferroni’s multiple-comparison test. ***P < 0.001, ****P < 0.0001 for comparison of F508del and WT to KO; ^#P < 0.05, ^###P < 0.001, ^####P < 0.0001 for comparison of F508del LUM/IVA treatment to F508del vehicle (DMSO) treatment.

Figure 5. In utero and postnatal LUM/IVA treatment reduces meconium ileus and improves the survival of CFTR-F508del newborn ferrets.

(A) Representative images of the intestine from newborn kits with the indicated homozygous genotypes. Arrow marks the intestinal obstruction in the CFTR-F508del kit with meconium ileus (MI). (B) Percentage of newborn homozygous CFTR-F508del kits with MI at birth with the indicated in utero treatment conditions. Pregnant jills harboring CFTR-F508del kits were untreated or treated with IVA or LUM/IVA beginning on embryonic day 28 (E28). Numbers in parentheses represent the number of births for each treatment condition evaluated. (C) One-week survival rates of homozygous CFTR-F508del kits that passed stool at birth for the indicated in utero and postnatal treatment conditions. In B and C, Fisher’s exact test was used to compare the untreated group to each of the treatment groups (NS, P = 0.154, *P = 0.0265, ****P < 0.0001) or the IVA-treated to the LUM/IVA–treated groups (^####P < 0.0001).

Figure 6. Pancreatic pathology of juvenile and adult WT and CFTR-F508del ferrets.

(A–C) H&E-stained sections of the pancreas from 2-month-old (A) WT and (B and C) 2 homozygous CFTR-F508del ferrets. These CFTR-F508del ferrets received only IVA in utero and postnatally until sacrifice at 2 months of age. Note the marked pancreatic fibrosis (*), acinar loss, cystic dilations (closed arrows), ductal dilatation (†), and pockets of inflammation (open arrows). Islet morphology was relatively intact, as shown in the boxed regions of the main panels and enlarged insets (a’–c’). Residual acinar cells with cystic dilatations are shown in the boxed regions of the main panels and enlarged insets (a”–c”). (D–F) Sections of the pancreas from 2 different adult CFTR-F508del ferrets that were reared on only IVA, later removed, and euthanized at more than 1 year of age. Insets show enlarged boxed regions in the main panels (d, e, f’, and f’’). Note the marked adipose replacement (†), ductal dilatation (*), and aggregated islets in pockets of fibrosis (arrows). (E and e) Masson’s trichrome staining for fibrosis (blue). Scale bars: 100 μm (A–C), 50 μm (a–c), 1 mm (D), and 250 μm (d).

Figure 7. In utero and postnatal LUM/IVA treatment of CFTR-F508del ferrets partially protects from pancreatitis.

(A) Masson’s trichrome–stained section of the pancreas from a CFTR-F508del ferret treated in utero and postnatally with LUM/IVA until 45 days of age and sacrificed at 101 days of age. There is marked pancreatic fibrosis (blue staining), adipose replacement (†), duct plugging (#), and islet aggregation (arrows). (B and C) Masson’s trichrome– (B) and H&E-stained (C) sections of the pancreas from a CFTR-F508del ferret treated in utero and postnatally with LUM/IVA until sacrificed at 51 days of age. The H&E-stained section highlights a localized area with cystic dilatation of acini and inflammatory infiltrates (*) in the animal continuously treated with LUM/IVA. When quantified, this type of pathology was observed in 4.1% of the area of the entire pancreas. Islets are marked by arrows and have normal architecture. (D) H&E-stained section of the pancreas from the 223-day-old CFTR-F508del ferret continuously treated with LUM/IVA. This ferret had normal fecal elastase levels at 140 days (4,682 μg EL-1/g feces) and showed no areas of acinar loss in the head, body, and tail of the pancreas and normal islet architecture (arrow). (E–G) Islet organization shown by histochemical staining of pancreatic sections for insulin (brown) in adult (E) WT (500 days old), (F) CFTR-F508del (505 days old) removed from LUM/IVA at 117 days, and (G) CFTR-F508del (223 days old) reared on LUM/IVA until euthanized. (H–J) Pancreatic sections stained with (H) H&E, (I) for insulin, and (J) for insulin with Masson’s trichrome from a CFTR-F508del ferret treated in utero and postnatally with LUM/IVA until 45 days of age and euthanized at 296 days of age. This ferret had undetectable fecal elastase levels before the necropsy. Aggregated islets are marked by an arrow and † marks fat. (a–i) Higher-power magnification images of the boxed regions in the main panels (A–I). Scale bars: 1 mm (A, B, D, H, and J), 500 μm (C, E–G, and I), 250 μm (a–d and h), and 100 μm (e–g and i).

Figure 8. Bacterial colonization of the lung in WT and CFTR-F508del ferrets.

(A) CFTR-F508del (FF) ferrets were treated in utero and postnatally with LUM/IVA (VX) until the indicated ages and bronchoalveolar lavage fluid (BALF) collected at the age indicated to assess bacterial load in the BALF and bacterial genus. The percentage of each bacterial genus is given on the left axis, with the color code given in the legend. The bacterial load is given on the right axis as colony-forming units (CFU) per mL. “ON” in A indicates that these 3 ferrets were treated with LUM/IVA for the entire time. (B and C) Longitudinal bronchoscopies of CFTR-F508del and WT ferrets assessing aerobic (B) and anaerobic (C) bacterial CFU in the BALF of independent ferrets shown in A. Significant differences were determined by 1-way ANOVA with Tukey’s multiple-comparison test. **P < 0.01, ****P < 0.0001 for comparison between WT and ON/OFF LUM/IVA CFTR-F508del ferret groups. NS, P > 0.05. Each animal underwent 1–4 bronchoscopies. (D–G) PAS-stained sections of inflated lungs from a CFTR-F508del ferret (D and E) reared on only IVA from E28 to 117 days of age, and then removed and sacrificed at 505 days of age and (F and G) reared on both IVA and LUM from E28 to 133 days of age, and then removed and sacrificed at 136 days of age due to acute health concerns. Mucus accumulation (asterisk) in the airways (d1, e, f1, and g) and submucosal glands (d2 and f2) are marked by asterisks and arrows. Boxed regions and arrow in the main panel point to areas of enlarged insets. Scale bars: 500 μm (D–G), 100 μm (d1, e, f1, and g), and 40 μm (d2 and f2).

Figure 9. CFTR-F508del ferrets have altered respiratory mechanics indicative of obstructive and restrictive lung disease.

FlexiVent pulmonary function testing was performed on 3 groups of ferrets: (i) CFTR-F508del ferrets (n = 14 measurements in 6 animals), (ii) WT ferrets that were paired controls in terms of age and sex (n = 14 measurements in 13 animals), and (iii) a larger cohort of WT ferrets (n = 160 measurements from a total of 68 animals). The average age of the paired cohort was 334 ± 49 days for CFTR-F508del and 388 ± 74 days for WT. The average age of the larger WT cohort was 328 ± 16 days. Data are shown as mean ± SEM and broken down by sex, with the number of measurements (n) used for each parameter. A mixed-effects model, which accounted for repeat measures, was used to compare genotypes and calculate P values. (A) Inspiratory capacity (IC) from age- and sex-matched pairs of WT and CFTR-F508del ferrets (P < 0.0001). (B) Percentage predicted IC (matched pairs, P = 0.0005; larger cohort, P = 0.0018). (C) Percentage predicted forced expiratory volume in 0.4 seconds (FEV_0.4) (matched pairs, P = 0.022; larger cohort, P = 0.0034). (D) Percentage predicted forced vital capacity (FVC) (matched pairs, P = 0.0061; larger cohort, P = 0.051). (E) FEV_0.4/FVC ratio (matched pairs, P = 0.076; larger cohort, P = 0.0023). (F) Pressure-volume loops (PV-loops) between pairs of WT and CFTR-F508del ferrets (P < 0.0001 for grouped analysis of both sexes). (G) Quasistatic compliance (Cst) (matched pairs, P = 0.0002; larger cohort, P < 0.0001). (H) Dynamic compliance (Crs) (matched pairs, P = 0.0001; larger cohort, P < 0.0001). (I) Resistance of the respiratory system (Rrs) (matched pairs, P = 0.126; larger cohort, P = 0.0046). The age and CFTR modulator parameters of F508del animals in the paired cohort can be found in the Supporting Data Values file for A.