Marcel Bijvelds

A functional CFTR assay using primary cystic fibrosis intestinal organoids

We recently established conditions allowing for long-term expansion of epithelial organoids from intestine, recapitulating essential features of the in vivo tissue architecture. Here we apply this technology to study primary intestinal organoids of people suffering from cystic fibrosis, a disease caused by mutations in CFTR, encoding cystic fibrosis transmembrane conductance regulator. Forskolin induces rapid swelling of organoids derived from healthy controls or wild-type mice, but this effect is strongly reduced in organoids of subjects with cystic fibrosis or in mice carrying the Cftr F508del mutation and is absent in Cftr-deficient organoids. This pattern is phenocopied by CFTR-specific inhibitors. Forskolin-induced swelling of in vitro–expanded human control and cystic fibrosis organoids corresponds quantitatively with forskolin-induced anion currents in freshly excised ex vivo rectal biopsies. Function of the CFTR F508del mutant protein is restored by incubation at low temperature, as well as by CFTR-restoring compounds. This relatively simple and robust assay will facilitate diagnosis, functional studies, drug development and personalized medicine approaches in cystic fibrosis.

Cystic fibrosis: Toward personalized therapies

Cystic fibrosis (CF), the most common, life-threatening monogenetic disease in Caucasians, is caused by mutations in the CFTR gene, encoding a cAMP- and cGMP-regulated epithelial chloride channel. Symptomatic therapies treating end-organ manifestations have increased the life expectancy of CF patients toward a mean of 40 years. The recent development of CFTR-targeted drugs that emerged from high-throughput screening and are capable of correcting the basic defect promises to transform the therapeutic landscape from a trial-and-error prescription to personalized medicine. This stratified approach is tailored to a specific functional class of mutations in CFTR, but can be refined further to an individual level by exploiting recent advances in ex vivo drug testing methods. These tests range from CFTR functional measurements in rectal biopsies donated by a CF patient to the use of patient-derived intestinal or pulmonary organoids. Such organoids may serve as an inexhaustible source of epithelial cells that can be stored in biobanks and allow medium- to high-throughput screening of CFTR activators, correctors and potentiators on the basis of a simple microscopic assay monitoring organoid swelling. Thus the recent breakthrough in stem cell biology allowing the culturing of mini-organs from individual patients is not only relevant for future stem cell therapy, but may also allow the preclinical testing of new drugs or combinations that are optimally suited for an individual patient.

Modeling rotavirus infection and antiviral therapy using primary intestinal organoids.

Despite the introduction of oral vaccines, rotavirus still kills over 450,000 children under five years of age annually. The absence of specific treatment prompts research aiming at further understanding of pathogenesis and the development of effective antiviral therapy, which in turn requires advanced experimental models. Given the intrinsic limitations of the classical rotavirus models using immortalized cell lines infected with laboratory-adapted strains in two dimensional cultures, our study aimed to model infection and antiviral therapy of both experimental and patient-derived rotavirus strains using three dimensional cultures of primary intestinal organoids. Intestinal epithelial organoids were successfully cultured from mouse or human gut tissues. These organoids recapitulate essential features of the in vivo tissue architecture, and are susceptible to rotavirus. Human organoids are more permissive to rotavirus infection, displaying an over 10,000-fold increase in genomic RNA following 24h of viral replication. Furthermore, infected organoids are capable of producing infectious rotavirus particles. Treatment of interferon-alpha or ribavirin inhibited viral replication in organoids of both species. Importantly, human organoids efficiently support the infection of patient-derived rotavirus strains and can be potentially harnessed for personalized evaluation of the efficacy of antiviral medications. Therefore, organoids provide a robust model system for studying rotavirus-host interactions and assessing antiviral medications.

rAAV-CFTRΔR Rescues the Cystic Fibrosis Phenotype in Human Intestinal Organoids and Cystic Fibrosis Mice

RATIONALE Gene therapy holds promise for a curative mutation-independent treatment applicable to all patients with cystic fibrosis (CF). The various viral vector-based clinical trials conducted in the past have demonstrated safety and tolerance of different vectors, but none have led to a clear and persistent clinical benefit. Recent clinical breakthroughs in recombinant adeno-associated viral vector (rAAV)-based gene therapy encouraged us to reexplore an rAAV approach for CF. OBJECTIVES We evaluated the preclinical potential of rAAV gene therapy for CF to restore chloride and fluid secretion in two complementary models: intestinal organoids derived from subjects with CF and a CF mouse model, an important milestone toward the development of a clinical rAAV candidate for CF gene therapy. METHODS We engineered an rAAV vector containing a truncated CF transmembrane conductance regulator (CFTRΔR) combined with a short promoter (CMV173) to ensure optimal gene expression. A rescue in chloride and fluid secretion after rAAV-CFTRΔR treatment was assessed by forskolin-induced swelling in CF transmembrane conductance regulator (CFTR)-deficient organoids and by nasal potential differences in ΔF508 mice. MEASUREMENTS AND MAIN RESULTS rAAV-CFTRΔR transduction of human CFTR-deficient organoids resulted in forskolin-induced swelling, indicating a restoration of CFTR function. Nasal potential differences demonstrated a clear response to low chloride and forskolin perfusion in most rAAV-CFTRΔR-treated CF mice. CONCLUSIONS Our study provides robust evidence that rAAV-mediated gene transfer of a truncated CFTR functionally rescues the CF phenotype across the nasal mucosa of CF mice and in patient-derived organoids. These results underscore the clinical potential of rAAV-CFTRΔR in offering a cure for all patients with CF in the future.

Cholesterol depletion and genistein as tools to promote F508delCFTR retention at the plasma membrane.

Background/aims: F508delCFTR-, but not wtCFTR-, expressing fibroblasts resemble Niemann Pick type C cells in the massive intracellular accumulation of free cholesterol. The recruitment and activation of F508delCFTR by cholesterol depletion was studied. Methods: Filipin staining, forskolin-stimulated anion efflux and FITC-dextran uptake were studied in control cells and fibroblasts treated with 2-hydroxypropyl β-cyclodextrin phosphatidylcholine large unilamellar vesicles to deplete cellular free cholesterol.Results: Treatment of F508delCFTR-, but not wtCFTR-, expressing fibroblasts with 2-hydroxypropyl β-cyclodextrin resulted in a reduction in cellular cholesterol and a potentiation of the forskolin-induced anion efflux. In addition, forskolin also promoted a massive increase in the rate of endocytosis in F508delCFTR fibroblasts, which was absent in genistein- or cyclodextrin-treated cultures.Conclusion: The results not only suggest that reducing cellular cholesterol may serve as pharmacotherapeutic tool in the treatment of cystic fibrosis but also reveal a novel mechanism for genistein regulation of F508delCFTR, i.e. retention by inhibition of endocytosis.

Guanylin and uroguanylin are produced by mouse intestinal epithelial cells of columnar and secretory lineage

Guanylin (GN) and uroguanylin (UGN), through activation of guanylyl cyclase C (GCC), serve to control intestinal fluid homeostasis. Both peptides are produced in the intestinal epithelium, but their cellular origin has not been fully charted. Using quantitative PCR and an improved in situ hybridization technique (RNAscope), we have assessed the expression of GN (Guca2a), UGN (Guca2b), and GCC (Gucy2c) in mouse intestine. In the crypts of Lieberkühn, expression of Guca2a and Guca2b was restricted to cells of secretory lineage, at the crypt’s base, and to a region above, previously identified as a common origin of cellular differentiation. In this compartment, comparatively uniform levels of Guca2a and Guca2b expression were observed throughout the length of the gut. In contrast, Guca2a and Guca2b expression in the villus–surface region was more variable, and reflected the distinct, but overlapping expression pattern observed previously. Accordingly, in jejunum and ileum, Guca2a and Guca2b were abundantly expressed by enterocytes, whereas in colon only Guca2a transcript was found in the surface region. In duodenum, only low levels of Guca2b transcript were observed in columnar cells, and Guca2a expression was restricted entirely to cells of the secretory lineage. Gucy2c was shown to be expressed relatively uniformly along the rostrocaudal and crypt–villus axes and was also found in the duodenal glands. Our study reveals novel aspects of the cellular localization of the GCC signaling axis that, apart from its role in the regulation of fluid balance, link it to pH regulation, cell cycle control, and host defense.

Effect of antibiotic treatment on fat absorption in mice with cystic fibrosis

Introduction:Improving fat absorption remains a challenge in cystic fibrosis (CF). Antibiotics (AB) treatment has been shown to improve body weight in CF mice. The mechanism may include improvement in fat absorption. We aimed to determine the effect of AB on fat absorption in two CF mouse models.Results:AB did not improve total fat absorption. Interestingly, AB accelerated the absorption of isotope-labeled fats, in both Δ/Δ and WT mice. The changes observed were not related to the solubilization capacity of bile or to changes in the bacteria in the small intestine. AB reduced the fecal excretion of cholate by ~50% (P < 0.05) in both CF mouse models, indicating improved intestinal bile salt absorption.Discussion:In conclusion, AB treatment does not improve total fat absorption in CF mice but does decrease fecal loss of bile salts and accelerate long-chain fatty acid (LCFA) absorption.Methods:For 3 weeks, we administered oral AB (ciprofloxacin/metronidazole) or control treatment to homozygous ΔF508 (Δ/Δ), cystic fibrosis transmembrane conductance regulator (CFTR) knockout (−/−), and wild-type (WT) mice and quantified fat absorption using a 72-h fat balance test. In Δ/Δ mice, we assessed fat absorption kinetics by administering tri-1-13C-palmitin and 1-13C-stearate intragastrically and determining the appearance of stable isotope-labeled fats in plasma. We quantified biliary and fecal bile salts (gas chromatography) and small intestinal bacteria (quantitative-PCR).

Ursodeoxycholate modulates bile flow and bile salt pool independently from the cystic fibrosis transmembrane regulator (Cftr) in mice

Cystic fibrosis liver disease (CFLD) is treated with ursodeoxycholate (UDCA). Our aim was to evaluate, in cystic fibrosis transmembrane regulator knockout (Cftr(-/-)) mice and wild-type controls, whether the supposed therapeutic action of UDCA is mediated via choleretic activity or effects on bile salt metabolism. Cftr(-/-) mice and controls, under general anesthesia, were intravenously infused with tauroursodeoxycholate (TUDCA) in increasing dosage or were fed either standard or UDCA-enriched chow (0.5% wt/wt) for 3 wk. Bile flow and bile composition were characterized. In chow-fed mice, we analyzed bile salt synthesis and pool size of cholate (CA). In both Cftr(-/-) and controls intravenous TUDCA stimulated bile flow by ∼250% and dietary UDCA by ∼500%, compared with untreated animals (P < 0.05). In non-UDCA-treated Cftr(-/-) mice, the proportion of CA in bile was higher compared with that in controls (61 ± 4 vs. 46 ± 4%; P < 0.05), accompanied by an increased CA synthesis [16 ± 1 vs. 10 ± 2 μmol·h(-1)·100 g body wt (BW)(-1); P < 0.05] and CA pool size (28 ± 3 vs. 19 ± 1 μmol/100 g BW; P < 0.05). In both Cftr(-/-) and controls, UDCA treatment drastically reduced the proportion of CA in bile below 5% and diminished CA synthesis (2.3 ± 0.3 vs. 2.2 ± 0.4 μmol·day(-1)·100 g BW(-1); nonsignificant) and CA pool size (3.6 ± 0.6 vs. 1.5 ± 0.3 μmol/100 g BW; P < 0.05). Acute TUDCA infusion and chronic UDCA treatment both stimulate bile flow in cystic fibrosis conditions independently from Cftr function. Chronic UDCA treatment reduces the hydrophobicity of the bile salt pool in Cftr(-/-) mice. These results support a potential beneficial effect of UDCA on bile flow and bile salt metabolism in cystic fibrosis conditions.

Cholic Acid Induces a Cftr Dependent Biliary Secretion and Liver Growth Response in Mice

The cause of Cystic fibrosis liver disease (CFLD), is unknown. It is well recognized that hepatic exposure to hydrophobic bile salts is associated with the development of liver disease. For this reason, we hypothesize that, CFTR dependent variations, in the hepatic handling of hydrophobic bile salts, are related to the development CFLD. To test our hypothesis we studied, in Cftr-/- and control mice, bile production, bile composition and liver pathology, in normal feeding condition and during cholate exposure, either acute (intravenous) or chronic (three weeks via the diet). In Cftr-/- and control mice the basal bile production was comparable. Intravenous taurocholate increased bile production to the same extent in Cftr-/- and control mice. However, chronic cholate exposure increased the bile flow significantly less in Cftr-/- mice than in controls, together with significantly higher biliary bile salt concentration in Cftr-/- mice. Prolonged cholate exposure, however, did not induce CFLD like pathology in Cftr-/- mice. Chronic cholate exposure did induce a significant increase in liver mass in controls that was absent in Cftr-/- mice. Chronic cholate administration induces a cystic fibrosis-specific hepatobiliary phenotype, including changes in bile composition. These changes could not be associated with CFLD like pathological changes in CF mouse livers. However, chronic cholate administration induces liver growth in controls that is absent in Cftr-/- mice. Our findings point to an impaired adaptive homeotrophic liver response to prolonged hydrophobic bile salt exposure in CF conditions.

Bile acid handling in cystic fibrosis: marked phenotypic differences between mouse models.

Dear Sir: Debray et al reported an intriguing study centered on the observation of delayed gallbladder emptying in cystic fibrosis (CF) mice, that is, mice with a genetically induced defect in CF transmembrane conductance regulator (CFTR) function.1 The authors propose that in CF mice, defective gallbladder emptying provokes cholecystohepatic shunting, namely, the (re)absorption of bile acids (BAs) from the gallbladder and transfer back to the canaliculi. It is hypothesized that, as a result, CF mice secrete less BA into the intestine, and that enterohepatic cycling of BAs is reduced. A low duodenal BA input may contribute importantly to the pathophysiology of CF, because it is likely to result in defective micellar solubilization of long-chain fatty acids. Fat malabsorption is a well-recognized symptom of CF, and we have shown that in CF mice, it is associated not only with impaired lipolysis, but also aberrant processing of long-chain fatty acids.2 Thus, Debray et al may ave uncovered a cardinal factor in the multifactorial tiology of CF-related fat malabsorption. Perhaps surprisingly, Debray et al show that the level of ecal BA excretion in Cftr / mice is comparable with noral mice. At variance with the present investigation, we and thers have previously shown that fecal BA output was ignificantly increased, both in mice homozygous for the ommon F508del-Cftr mutation and in Cftr / animals, in omparison with normal littermates.2,3 In CF, enhanced feal BA excretion may result from a reduced capacity for ileal A absorption.4 The terminal ileum is critical for reclaiming BAs, as it actively absorbs BAs through the apical sodiumdependent BA transporter (ASBT). Consequently, the normal level of fecal BA excretion observed in Cftr / mice by Debray et al may result from a low duodenal influx that is offset by a low absorption rate. The dramatic reduction in Slc10a2 (encoding ASBT) expression and ASBT protein abundance observed in the ileum of these particular Cftr / mice seems to fit in with this model. However, it is not generally observed in other CF mouse models. Previously, we observed robust ASBT immunostaining in the ileum of homozygous F508del-Cftr and Cftr / animals.5 In fact, ASBT partitioning along the crypt–villus axis was indistinguishable from normal animals. Immunostaining was observed only in the distal ileum, concordant with functional assays, which showed that ileal taurocholate uptake was unaffected in homozygous F508del-Cftr mice and only marginally reduced (17%) in Cftr / anials compared with wild-type littermates. These observations have since been corroborated by reudenberg et al, who reported an increase in fecal BA xcretion in a different F508del-Cftr model, and a normal SBT abundance in the ileum of these mice.3 Accordingly, Hardcastle et al have shown that Na -coupled taurocholate uptake in ileal brush-border membrane vesicles of homozygous F508del-Cftr mice was unperturbed, strongly arguing against a lowered ASBT activity.6 Debray et al propose that Slc10a2 expression in the ileum of Cftr / mice is down-regulated as a result of lowered BA exposure (and uptake). However, whether defective gallbladder emptying significantly reduces intestinal BA levels in these animals is speculative, because the duodenal BA influx was not directly assessed. Remarkably, it was observed that Cftr / mice, despite the very low levels of intestinal ASBT, had normal levels of BA in the portal blood, which does not support this model. This thesis is also difficult to reconcile with earlier data showing that luminal BA depletion (by cholestyramine feeding) increased Slc10a2 expression in mouse ileum,7 and enhanced BA absorption in guinea pigs,8 indicating that As exert a negative feedback control over Slc10a2 expression. In conclusion, defective gallbladder emptying, leading to ncreased cholecystohepatic shunting of BAs, associated with educed ASBT expression in terminal ileum, but not in allbladder, may explain most of the abnormalities in BA etabolism and transport in the CF mice studied by the ousset laboratory. However, such abnormalities are not ound uniformly in other CF mouse models. This realization nderscores the need for additional studies exploring the ause of these differences in BA handling between separate F mouse models, including possible effects of genetic backround, the intestinal microbiome, and diet composition eg, the use of a laxative in the drinking water to prevent ntestinal obstruction in CF mice; applied in studies 1 and 3 ut not in 2 or 5). Such studies may glean important inights into the disease mechanism of CF, and the causes of isease variability. Awaiting the outcome, caution should be rged in extrapolating the results of a particular CF mouse tudy on BA transport and metabolism to human CF.