Journal of Cystic Fibrosis Volume 10 Suppl 2 (2011) S152–S171
Mouse models of cystic fibrosis: ...
M. Wilke et al. / Journal of Cystic Fibrosis Volume 10 Suppl 2 (2011) S152–S171 S153
great progress from these models, the...
S154 M. Wilke et al. / Journal of Cystic Fibrosis Volume 10 Suppl 2 (2011) S152–S171
Table 1
CFTR mutant mice
Mouse Mutati...
M. Wilke et al. / Journal of Cystic Fibrosis Volume 10 Suppl 2 (2011) S152–S171 S155
Table 2
Comparative CF pathology in h...
S156 M. Wilke et al. / Journal of Cystic Fibrosis Volume 10 Suppl 2 (2011) S152–S171
deficiency determine disease severity,...
M. Wilke et al. / Journal of Cystic Fibrosis Volume 10 Suppl 2 (2011) S152–S171 S157
Fig. 1. Goblet cells in CF mutant mou...
S158 M. Wilke et al. / Journal of Cystic Fibrosis Volume 10 Suppl 2 (2011) S152–S171
Fig. 2. Immunohistological analysis o...
M. Wilke et al. / Journal of Cystic Fibrosis Volume 10 Suppl 2 (2011) S152–S171 S159
Fig. 4. Isoproterenol-stimulated sali...
S160 M. Wilke et al. / Journal of Cystic Fibrosis Volume 10 Suppl 2 (2011) S152–S171
4.8. Abnormal bone and cartilage form...
M. Wilke et al. / Journal of Cystic Fibrosis Volume 10 Suppl 2 (2011) S152–S171 S161
Fig. 6. CFTR immunostaining in mouse ...
S162 M. Wilke et al. / Journal of Cystic Fibrosis Volume 10 Suppl 2 (2011) S152–S171
Fig. 7. Summary of short-circuit curr...
M. Wilke et al. / Journal of Cystic Fibrosis Volume 10 Suppl 2 (2011) S152–S171 S163
mouse lung and sensitive to the CFTR ...
S164 M. Wilke et al. / Journal of Cystic Fibrosis Volume 10 Suppl 2 (2011) S152–S171
quickly depending on the dose and met...
M. Wilke et al. / Journal of Cystic Fibrosis Volume 10 Suppl 2 (2011) S152–S171 S165
to be determined. It is hypothesized ...
S166 M. Wilke et al. / Journal of Cystic Fibrosis Volume 10 Suppl 2 (2011) S152–S171
7.3. Anti-inflammatory drugs
In additi...
M. Wilke et al. / Journal of Cystic Fibrosis Volume 10 Suppl 2 (2011) S152–S171 S167
bronchitis and emphysema in β-epithel...
S168 M. Wilke et al. / Journal of Cystic Fibrosis Volume 10 Suppl 2 (2011) S152–S171
modifier gene for cystic fibrosis lung ...
M. Wilke et al. / Journal of Cystic Fibrosis Volume 10 Suppl 2 (2011) S152–S171 S169
duce increased expression of neutroph...
1 s2.0-s1569199311600209-main
1 s2.0-s1569199311600209-main
Upcoming SlideShare
Loading in …5

1 s2.0-s1569199311600209-main


Published on

  • Be the first to comment

  • Be the first to like this

No Downloads
Total views
On SlideShare
From Embeds
Number of Embeds
Embeds 0
No embeds

No notes for slide

1 s2.0-s1569199311600209-main

  1. 1. Journal of Cystic Fibrosis Volume 10 Suppl 2 (2011) S152–S171 Mouse models of cystic fibrosis: Phenotypic analysis and research applications Martina Wilkea,1 , Ruvalic M. Buijs-Offermanb , Jamil Aarbioub,2 , William H. Colledgec , David N. Sheppardd , Lhousseine Touquie , Alice Bota , Huub Jornaa , Hugo R. de Jongea , Bob J. Scholteb, * a Erasmus MC, Biochemistry Department, 3000 CA Rotterdam, The Netherlands b Erasmus MC, Cell Biology Department, 3000 CA Rotterdam, The Netherlands c Physiological Laboratory, Department of Physiology, Development and Neuroscience, University of Cambridge, Cambridge CB2 3EG, UK d University of Bristol, School of Physiology & Pharmacology, Medical Sciences Building, University Walk, Bristol BS8 1TD, UK e Unité de Défense Innée et Inflammation, Unité Inserm U.874, Institut Pasteur, 75015 Paris, France Abstract Genetically modified mice have been studied for more than fifteen years as models of cystic fibrosis (CF). The large amount of experimental data generated illuminates the complex multi-organ pathology of CF and raises new questions relevant to human disease. CF mice have also been used to test experimental therapies prior to clinical trials. This review recapitulates the major phenotypic traits of CF mice and highlights important new findings including aberrant alveolar macrophages, bone and cartilage abnormalities and abnormal bioactive lipid metabolism. Novel data are presented on the intestinal and nasal physiology of F508del-CFTR CF mice backcrossed onto different genetic backgrounds. Caveats, and sources of variability including age, gender and animal husbandry, are discussed. Interspecies differences limit comparison of lung pathology in CF mice to the human disease. The recent development of genetically modified pigs and ferrets heralds the application of more advanced animal models to CF research and drug development. © 2011 European Cystic Fibrosis Society. Published by Elsevier B.V. All rights reserved. Keywords: Cystic fibrosis; Animal models; CFTR mice; Lung disease; Intestinal disease; F508del-CFTR 1. Introduction The life expectancy of cystic fibrosis (CF) patients in- creased from around 20 to 40 years over the past two decades, mainly due to improved nutrition and treatment of chronic lung disease. However, there is still an urgent unmet need to improve the lifespan and quality of life of CF patients. The complex multi-organ pathophysiology of secretory epithelia in CF is not yet completely understood, and a treatment that sufficiently corrects the primary defect remains to be developed. * Corresponding author: B.J. Scholte, PhD, Erasmus MC, Cell Biology Department, PO Box 2040, 3000 CA Rotterdam, The Netherlands. Tel.: +31 10 7043205; fax: +31 10 7044743. E-mail address: (B.J. Scholte) 1 Present address: Erasmus MC, Clinical Genetics Department. 2 Present address: Galapagos NV, Leiden. 1569-1993/$ - see front matter © 2011 European Cystic Fibrosis Society. Published by Elsevier B.V. All rights reserved. Because access to relevant patient tissues is limited, genet- ically modified mice are widely used to study CF pathology and experimental therapeutics. These models have proven to be very useful, but they have important limitations, due to obvious physiological differences between humans and mice [1–3]. Recently, pig [4,5] and ferret [6] cystic fibrosis trans- membrane conductance regulator (CFTR) knockout animals were reported, opening an exciting new field. Both models show dominant intestinal disease [6,7], with severe perinatal lethality. There is also evidence for reduced clearance of bacterial lung infection in a limited number of long-lived CF pigs [8]. In both models, F508del-CFTR mutants are being developed, which will facilitate the testing of novel therapeu- tic strategies for CF. The challenge for the time to come will be to manage intestinal disease in CF pigs and ferrets to allow better understanding of early lung disease, which is still an underexplored area of CF pathology. While we can expect
  2. 2. M. Wilke et al. / Journal of Cystic Fibrosis Volume 10 Suppl 2 (2011) S152–S171 S153 great progress from these models, their application is limited by the considerable practical and financial issues involved. This paper will focus on CF mouse models, presenting an update on new developments and novel data, in particular on the properties of the mutant strains made available to the CF community through the EuroCareCF project. 2. CF mutant mice, portrait of an unruly family Shortly after the discovery of the CFTR gene, several groups set out to generate mouse strains with mutations in the Cftr locus using the homologous recombination technology and mouse embryonic stem cells pioneered by Mario Capec- chi, Martin Evans and Oliver Smithies who jointly received the Nobel Prize for these innovations in 2007. Table 1 lists the genetically modified mouse strains produced by different groups. Details of mouse strain nomenclature and availability can be found at the Mouse Genome Informatics website ( Table 2 summarizes briefly the major phenotypic traits of CF mice in comparison with the human disease. Several loss of function, or “knock-out” models were made by interrupting the mouse Cftr coding region with a selectable marker. The Cftrtm1Hgu shows low level of residual activity of wild-type CFTR and lower mortality, compared to the other two frequently used knockout strains Cftrtm1Cam and Cftrtm1Unc (Table 1). In addition, common CF causing mutations were intro- duced into the mouse CFTR amino acid sequence. Three strains with the F508del mutation in the mouse Cftr gene were generated, two of which carry an intronic insertion that may interfere with gene transcription (Table 1). By contrast, the Cftrtm1Eur allele was made by the two step “hit and run” recombination procedure, which removes the selectable marker. Therefore, this allele only contains the F508del mu- tation and has normal transcription rates in all tissues [9,10]. All three strains have been successfully used in phenotypic studies. The F508del mutation in the mouse Cftr context causes severely reduced amounts of fully glycosylated CFTR protein (band C) similar to the human form, and a phenotype comparable to CFTR loss of function mutants [9]. In cultured gallbladder epithelial cells from F508del mutant mice, the number of active CFTR Cl− channels observed in patch- clamp experiments is increased by incubating cells at low temperature [9], suggesting a similar temperature-dependent processing defect to the F508del mutation in human CFTR. Moreover, the G551D mutation which disrupts channel gating has the typical CF phenotype caused by loss of function mutants [11]. The less severe phenotype of the F508del- and G551D-CFTR models when compared to CFTR knock- outs, particularly lower perinatal mortality due to intestinal obstruction, has been attributed to residual CFTR activity. To overcome the limitations imposed by the severe in- testinal disease presented by many knockout strains, “gut corrected” hybrid strains were created using a transgene ex- pressing CFTR from a promoter specific for the intestinal epithelium (Table 1). The most recent and important addition to the CF mouse family is a conditional (“loxed”) Cftr deletion that can be created by cell specific expression of the cre recombinase [12]. This genetically modified mouse will allow the contribution of different cell lineages to CF pathology to be elucidated. Although not a CFTR mouse model, genetically modi- fied mice overexpressing the β subunit of the epithelial Na+ channel (ENaC) are important because they possess a lung phenotype resembling that of CF [13–16]. For further infor- mation about this mouse model and its applications, see Zhou et al. [17]. 3. Mouse models, sources of variation and practical considerations CF mutant mice are generally much less fertile than normal mice, which makes it difficult to breed them from homozy- gous mutant pairs. This is partly due to frequent obstruction of the male vas deferens [18,19]. However, homozygous fe- males are also poor breeders. Occasionally mice can be bred successfully from homozygous pairs. However, there is risk of “genetic drift”, the selection of compensating mutations in either the background or the Cftr allele, which affect pheno- type in an unpredictable way. Breeding CF mutant mice from heterozygous animals is therefore recommended strongly. Im- portantly, this breeding strategy yields age- and sex-matched normal littermates, which are mandatory control for most experiments. Breeding requires considerable investment in DNA genotyping of all neonates by ear or tail clips, the identification of individual animals by chip, toe- or ear-clips and a database to track individual animals. Genotyping is primarily required to identify heterozygous breeders, because homozygous mutants are easily recognized by their typical white incisor teeth [20]. There is consensus about the most prominent phenotypic abnormalities of different Cftr mutant mice. Yet, close com- parison of published data from different laboratories shows considerable variation and even controversy about the nature and extent of lung disease. One obvious source of variation is the different Cftr mutations, some of which are associated with residual CFTR function (Table 1). Furthermore, Cftr alleles have been bred onto different genetic backgrounds, which may cause subtle differences in phenotype, presumably because of the expression of different “modifier genes”. Com- plete genomes of several laboratory mouse strains are now available (for further information, see Sanger mouse genome project, This resource will provide a new and powerful tool to ana- lyze phenotypic differences between CFTR mutant strains. It will also facilitate comparative studies of modifier genes and potential therapeutic targets in the human population. Animal husbandry is a critical issue, especially when in- vestigating innate adaptive responses (e.g. lung inflammation), which are by their very nature designed to reflect external conditions. This can cause differences between strains, con- tribute to inter-individual variation, and lead to differences in results between laboratories. Unfortunately, there is no
  3. 3. S154 M. Wilke et al. / Journal of Cystic Fibrosis Volume 10 Suppl 2 (2011) S152–S171 Table 1 CFTR mutant mice Mouse Mutation Cftr mRNA Genetic Survival to Body wt Contact References background maturity Null mutations Cftrtm1Unc S489X Ex10 R Not detectable* C57Bl/6 <5% 10–25% reduction BH Koller/Jax Labs [113,158] Cftrtm1Cam R487X Ex10 R Not detectable 129S6/Sv/Ev <5% 20% reduction WH Colledge [159] Cftrtm1Hsc M1X Ex1 R Not detectable CD1 x 129 25% Delayed LC Tsui [160] Cftrtm3Bay Ex2 R Not detectable C57Bl/6 x 129 40% Reduced AT Beaudet [161] Cftrtm3Uth Y122X Ex4 R Not detectable C57Bl/6 25% 25–50% reduction M Capecchi/PB Davis [113,162] Hypomorphic mutations Cftrtm1Hgu ** Ex10 I 10% of wt MF1 x 129 90% No reduction J Dorin [113] Cftrtm1Bay Ex3 I <2% wt C57Bl/6 x 129 40% 70% reduction AL Beaudet [163] F508del mutations Cftrtm2Cam F508del R 30% of wt 129S6/Sv/Ev <5% 10–20% reduction WH Colledge [164] Cftrtm1Kth F508del R Low in intestine C57Bl/6 x 129 40% 10–50% reduction KR Thomas/Jax labs [165] Cftrtm1Eur F508del (H&R) Normal levels FVB/129; FVB 90% 10–20% reduction BJ Scholte [9] C57Bl/6 Other point mutations Cftrtm2Hgu G551D R 50% of wt CD1/129 65% 30–50% reduction J Dorin [11] Cftrtm3Hgu G480C (H&R) Normal levels C57Bl/6/129 Normal No reduction J Dorin [166] Cftrtm2Uth R117H R 5–20% of wt C57/Bl6 95% 10–25% reduction M Capecchi/PB Davis [113,162] Transgenes Mouse Transgene Promoter Expression Phenotype References Tg(FABPCFTR) CFTR Rat intestinal fatty acid Intestinal villus epithelia Rescue of CF intestinal pathology [167] binding protein Tg(CCSPScnn1b) Scnn1b Clara cell secretory Airway surface epithelia Na+ hyperabsorption [13] protein (CCSP) Reduced airway surface fluid volume Mucus accumulation, CF-like lung disease; 40% survival. Legend: Mutations are introduced by homologous recombination in embryo stem cells. Key: R, replacement; I, insertion; (H&R), Hit & Run; * a splice variant was detected [168]. Transgenes are created by injection into mouse oocytes. General rules on mouse gene and strain nomenclature can be found at the Mouse Genome Informatics guidelines website. ** The strain Cftrtm1Hgu is also referred to as CftrTgH(neoim)Hgu due to the unique nature of this recombinant which produces low level normal CFTR due to a splice variant introduced by the recombination event. global standard protocol for animal husbandry. Laboratories therefore differ in the level of pathogen exposure. The Rot- terdam animal facility maintains a FELASA+ pathogen-free standard using sterilized ventilated cages supplied with steril- ized, acidified water. Separation of animals causes stress and is avoided whenever possible; nesting material (e.g. sterile paper tissue) is always provided. Strict light and dark regimes are maintained; nursing females are not disturbed until ten days after birth to reduce postnatal mortality. CF homozygous mutant animals do not travel very well and often die or suffer excessive weight loss during long distance transport. Therefore, animals are moved at least two weeks in advance of experiments to allow them to adapt. Under these condi- tions we can maintain colonies of Cftrtm1Eur F508del and Cftrtm1Cam knockout mice in different genetic backgrounds on normal chow at less than 20% mortality for homozygous mutants. Gender is another source of variation that should not be underestimated. Male and female mice differ with respect to many physiological responses relevant to CF studies. This is well illustrated in a recent study of a mouse model of asthma [21]. Gender differences should be taken into account when designing experiments and analyzing data. Finally, age is another important variable. Its importance is demonstrated by the progressive nature of lung disease in several CF mouse strains. Together, these considerations place severe demands on the size and selection of experimental groups necessary to generate reproducible data and statistical power. 4. CF mouse phenotype The major phenotypic abnormalities of CF mutant mice are briefly summarized in Table 2. This brief overview does not do justice to all the available literature. In the mutant strains investigated, pancreatic pathology is not a prominent feature, although it was reported in a C57Bl/6 Cftr knockout backcross [19]. Although pancreatic insufficiency correlates with severe pathology, it is not seen in all CF patients. Similarly, male infertility caused by obstruction or absence of the vas deferens is frequent in CF mice depending on the strain and genetic background [18,19]. However, this pathology appears less severe than in humans where absence of the vas deferens is observed in all male CF patients. Apparently, factors other than CFTR
  4. 4. M. Wilke et al. / Journal of Cystic Fibrosis Volume 10 Suppl 2 (2011) S152–S171 S155 Table 2 Comparative CF pathology in human and mouse CF mutants Human pathology [27,169] Tissues affected in CFTR mutant Mouse pathology [1,2] Recurrent bacterial lung infections, chronic inflammation, airway remodeling, irreversible loss of function [170] Lung No CF type obstructive lung disease. Inflammation and tissue remodeling in aging animals. Hyper-inflammation reduced clearance upon challenge with bacteria or LPS. Pro: [19,105,114, 117,171,172] Con: [173] Hyposecretion, viscous mucus, plugging [174,175] Submucosal gland (nasal, bronchial) Nasal and tracheal glands present, reduced cAMP-induced secretion in bronchial glands of mutant mice [81,82] Bacterial biofilm, viscous/purulent mucus, defective mucociliary clearance, goblet cell hyperplasia, inflammation, tissue remodeling [176,177] Trachea, bronchial epithelium (ciliated/ goblet) Reduced mucociliary clearance (contested). Electrophysiology influenced by CaCC (TMEM16A). Pro [84] Con: [83] Distal airways involved in early disease, CFTR expression in Clara cells [95] Broncheoli (distal epithelium, Clara cells) Spontaneous progressive remodeling and inflammation (strain- and age-dependent) [19,102,172,178] No known pathology, CFTR expression in cultured type II cells Alveoli (type I/II) Cftr expression in type II cells, abnormal fluid secretion in mutant mice. [97,99]. Meconium ileus, chronic malabsorption, reduced body weight, failure to thrive (pancreas deficiency), typical electrophysiology. [31,179, 180] Intestine Lethal mucus plugging, lipid and bile acid reabsorption defect, reduced body weight, failure to thrive. Goblet cell hyperplasia in ileum, mucus accumulation in intestinal crypts, meconium ileus equivalent, chronic inflammation, abnormal electrophyiology. [1,2,25,60,181, 182] Frequent biliary disease, progressive cirrhosis [43,183– 185] Liver/Biliary duct No morbidity, progressive liver disease in aging CF KO mice biliary duct damage in KO mice upon induced colitis [19,186,187] No frequent pathology [188] Kidney Functional CFTR in proximal tubule. In CF mice no overt pathology, abnormalities in Na+ and K+ transport, defective endocytosis. [41,42,189,190] Increased frequency of bile stones Gallbladder No cAMP-induced Cl− or water secretion [45,191] Pancreatitis, atrophy, digestive enzyme secretion deficiency, diabetes, abnormal bicarbonate secretion and mucous plugging of ducts [192,193] Pancreas Normal function and ductal fluid secretion, occasional duct plugging in Cftrtm1Unc mice [19,194] No pathology Tooth Abnormal enamel (white incisor) [20] Osteopenia, osteoporosis [73,74] Bone Abnormal bone density. [67–69] No pathology, reduced flow rate [195] Salivary gland Adrenergic fluid secretion defective [38] High NaCl in sweat (defective reabsorption duct) [169] Sweat gland Foot pad glands only, no report of deficiency. [196] Absent (CBAVD, male infertility 100%) Vas deferens Frequent plugging or absent (male infertility) [18,19] Frequent obstruction Oviduct Frequent obstruction (female infertility) No pathology reported Cornea Defective chloride permeability, enhanced amiloride response, attenuated uptake of Pseudomonas [197–199] Smooth muscle cell-related pathology in lung and diaphragm suspected but not firmly established Smooth muscle cell CFTR expression, altered function and innervation, degeneration in CF KO mice [47–50,200] No pathology reported Mast cells CFTR expression, enhanced presence in CF KO intestine, no evidence for cell autonomous abnormality [59,60] No pathology reported Macrophages CFTR expression, reduced capacity to kill Pseudomonas, abnormal proinflammatory cytokine secretion [51,55–57]
  5. 5. S156 M. Wilke et al. / Journal of Cystic Fibrosis Volume 10 Suppl 2 (2011) S152–S171 deficiency determine disease severity, in this case probably the expression of alternative pathways for Cl− transport. 4.1. Intestinal disease in CF mouse models Similar to humans, intestinal disease is prominent in all CF mutant mice. Depending on the strain and laboratory conditions many mutant mice may die of intestinal blockage around the time of weaning. This is a stochastic effect at- tributed to abnormal fluid and mucus secretion (meconium ileus equivalent). To reduce the frequency of intestinal ob- struction, mutant mice are often fed a liquid diet (Peptamen™) [22]. Because this drastic dietary intervention may affect the phenotype of both control and mutant mice, the use of a mild laxative (Poly Ethylene Glycol; PEG) in combination with normal chow is recommended as an alternative [23]. In the Rotterdam facility, Cftrtm1Eur and Cftrtm1Cam strains (FVB/n and C57Bl/6) survive well into adulthood (10–30 weeks, 80–90%) on normal chow based on animal fat (Hope Farm™ SRMA 2131, Abfarm, Woerden, the Netherlands) combined with 6% (w/v) PEG in the drinking water for the knockout mice. Male and female mutant mice of all strains show a mild, but significant (10% average), reduction of body weight compared to normal age- and sex-matched littermates. Although there is no evidence of an essential lipid deficiency [24], mutant mice display reduced lipid reabsorption, which is most pronounced in the Cftrtm1Cam strain [25]. Interestingly, the length and weight of the small intestine is larger in mutant Cftrtm1Eur C57Bl/6 compared to wild-type (10%, N = 18, P < 0.01, Wilke et al., unpublished data [26]). In the mouse and human intestine, different proximal to distal regions with distinct morphological and functional characteristics can be distinguished. In the duodenum, bile from the gallbladder and digestive enzyme secretions from the pancreas are delivered through a common duct and mixed with the acidic products of the stomach. In the jejunum and ileum, regulated fluid and mucus secretion maintain the optimal milieu required for lipid, sugar and amino-acid absorption. The colon, including the caecum, is involved in mucus and bicarbonate secretion, and net ion and fluid reabsorption, compacting the intestinal content to faeces. In addition to proximal to distal gradients along the length of the intestine, the epithelium is organized vertically by invaginations called crypts and finger-like protrusions called villi. Stem cells at the bottom of the crypts continuously produce new epithelial cells that differentiate to form the different cell types, lining crypts and villi. CFTR is expressed mainly in enterocytes where it is involved in chloride and bicarbonate secretion. The role of CFTR in intestinal ion transport [27,28] and in the small intestine [29,30] has been reviewed. The ion transport prop- erties of the colon, including the role of CFTR were reviewed by Kunzelmann and Mall [31]. Here, we present general and novel data relevant to this review on CF mouse models. Figure 1 demonstrates that the jejunum, ileum and colon of backcrossed Cftrtm1Eur C57Bl/6 and FVB/n mice exhibit a normal structure of crypts and villi. In the ileum a distinct hyperplasia of goblet cells is observed, comparable to published data for the original outbred strain (FVB/129) [10]. Consistent with this observation, the expression of the goblet cell marker Clca3/Gob5 is increased 3-fold in the intestine of CF mutant mice [32] and individual goblet cells appear larger than normal, possessing “plumes” of mucous material. Mucus plugs are often seen in the crypts and lumen. A recent study in Cftrtm1Kth F508del (C57Bl/6) mice highlights the crucial role of CFTR dependent bicarbonate secretion in PGE2-induced intestinal mucus secretion and hy- dration [33,34]. In the proposed model, bicarbonate secretion by CFTR in enterocytes is required for the proper expansion of tightly packed mucus-calcium ion complexes, which are secreted from vesicular granules by neighbouring goblet cells. In the absence of sufficient CFTR activity, the mucus does not unfold properly, which might play an important role in CF pathology. This indirect model for intestinal mucus secretion was invoked because CFTR expression in goblet cells has not been demonstrated directly, whereas enterocytes show relatively high levels of CFTR expression. If it can be shown that CFTR-mediated bicarbonate secretion plays a similar role in mucus release in human airway epithelia, this would have important implications for our understanding of CF lung disease. The mechanism causing the apparent goblet cell hyperpla- sia in the ileum of CF mutant mice has not been established. However, abnormalities of mucus composition in Cftrtm1Unc knockout mice [35], suggest an effect on goblet cell differen- tiation and function, possibly by an adaptive process. Immunohistochemistry using the R3195 anti-CFTR anti- body (CR Marino, University of Tennessee) shows CFTR expression on the apical membrane of enterocytes in villi and crypts of wild-type C57Bl/6 mice (Fig. 2). This signal is strongly reduced in Cftrtm1Eur F508del mutant mice and absent in Cftrtm1Cam knockout mice. Similar results were reported using the original Cftrtm1Eur 129/FVB strain [9] and the Cftr partial knockout Cftrtm1Hgu (also designated CftrTgh(neoim)Hgu) [36]. Consistent with these data, electro- physiological analyses demonstrate the absence or severe reduction of CFTR function in CF mutant mice. For exam- ple, Fig. 3 shows that in three different genetic backcrosses of the Cftrtm1Eur F508del allele forskolin-induced (cAMP- dependent) short-circuit current (Isc), a measure of CFTR- mediated transepithelial ion transport, is reduced by more than 80%. Figure 3 also reveals that the carbachol response (Ca2+-dependent) is reduced, which in ileal epithelia reflects the activity of Ca2+-activated K+ channels in the basolateral membrane that facilitate CFTR-mediated Cl− secretion. In CFTR knockout mice of the same genetic background, both the forskolin and carbachol responses are completely lacking, suggesting that the residual Isc response in intestinal epithelial of Cftrtm1Eur F508del mice reflects residual CFTR activity (De Jonge & Bot, unpublished). As for human CFTR, the intracellular transport defect and severely reduced activity of F508del-CFTR in Cftrtm1Eur homozygous mice is rescued by low temperature incubation [9]. Full restoration of wild-type CFTR protein processing
  6. 6. M. Wilke et al. / Journal of Cystic Fibrosis Volume 10 Suppl 2 (2011) S152–S171 S157 Fig. 1. Goblet cells in CF mutant mouse intestine. Comparison of representative paraffin sections from mouse jejunum, ileum and colon stained with Alcian blue to identify goblet cells. Wild-type C57Bl/6 (A) and FVB (C) are compared with their Cftrtm1Eur C57Bl/6 F508del (B) and Cftrtm1Eur FVB F508del (D) homozygous mutant littermates (12–16 weeks). Ileum derived from both mutant strains (B and D) show significant hypertrophy of goblet cells. No differences in jejunum and colon between wild-type and mutant mice are observed (original magnification 20×). and channel activity was observed in stripped ileal mucosa from Cftrtm1Eur homozygous mice after 12 hours incubation at 26°C [37] (Wilke, de Jonge, in preparation). Based on this observation, small molecules that rescue the cell surface expression of mutant CFTR (termed CFTR correctors) are now being tested in Cftrtm1Eur homozygous mice. 4.2. Salivary glands Best and Quinton [38] first demonstrated that β-adrenergic (cAMP)-stimulated fluid secretion by salivary glands is a non- invasive and highly discriminative assay of CFTR function. In this assay, atropine is first injected to prevent the stimula- tion of secretion by cholinergic agonists before isoproterenol is injected to activate β-adrenergic-stimulated fluid secretion. Subsequently, Noël et al. [39] demonstrated that this assay can be used to evaluate the efficacy with which small molecules restore function to Cftrtm1Eur F508del mutants in vivo. This assay offers the advantage of detecting very low levels of residual function in mutant mice (Fig. 4). Moreover, it can be performed repetitively in the same mouse, serving as a highly sensitive indicator of CFTR function [39]. Other aspects of the physiology of the salivary acini and glands in normal and CF mutant mice were described recently by Catalan et al. [40]. The authors’ data demonstrate that submandibular duct cells co-express CFTR and ENaC and that CFTR is required for effective reabsorption of NaCl [40]. 4.3. Kidney CFTR is expressed in the proximal tubules of the mouse kidney, localized to the apical and endosomal compartments of epithelial cells. Though no overt kidney pathology is present, abnormalities in urinary protein secretion were ob-
  7. 7. S158 M. Wilke et al. / Journal of Cystic Fibrosis Volume 10 Suppl 2 (2011) S152–S171 Fig. 2. Immunohistological analysis of CFTR expression in the ileum. CFTR protein was stained with the R3195 anti-Cftr antibody on paraffin sections. Intense CFTR staining (brown) is observed in the ileum of FVB/n wild-type mice at the apical border of crypt and villus epithelial cells (A). Staining in Cftrtm1Eur F508del FVB mutant littermates is much less intense and more diffuse (B). Tissue of the Cftrtm1Cam FVB knockout mouse serves as negative control (C). This figure is representative for all backcrossed strains described in the present paper (original magnification 20×). Fig. 3. Short-circuit measurements (Isc) of muscle-stripped excised ileum. Muscle stripped mucosa from ileum was analysed in an Ussing chamber as described in [77]. Bars represent average initial short-circuit current (Isc) or the change of Isc after sequential addition of the cAMP agonist forskolin (forsk), the Ca2+ agonist carbachol and glucose. Isc measurements were performed on ileum derived from the Cftrtm1Eur F508del strain in three different congenic backgrounds (F13) A: FVB/n (F13) B: C57Bl/6 (F13) C: 129/sv (F12) comparing homozygous mutant (dd, open bars) with homozygous normal (++, black bars). In all homozygous mutant Cftrtm1Eur F508del mice, forskolin responses are significantly lower than normal (P < 0.001), but residual activity is detected. Similar results are observed with carbachol. Glucose is added to test a CFTR-independent function and to evaluate tissue viability. All animals were adult males and females in the range of 13–30 weeks. Data presented are means ± SEM (n = number of mice, two pieces of ileum per mouse, FVB +/+, n = 9; FVB dd, n = 10; C57Bl/6 +/+, n = 10, C57Bl/6 d/d, n = 15; 129sv +/+, n = 11, 129sv d/d, n = 8). served in Cftrtm1Cam knockout mice, suggesting a role for CFTR in membrane trafficking and brush border matura- tion. To a lesser extent these abnormalities were also seen in Cftrtm2Cam and Cftrtm1Eur F508del mice, where they are attributed to partial processing of the mutant protein in this tissue [41,42]. 4.4. Liver CF associated liver disease (CFLD) is an increasing clinical problem, affecting more than 50% of adult CF patients. CFTR is expressed in cholangiocytes, where it is presumably involved in fluid homeostasis in the biliary duct system [43]. In CF mice, no obvious liver pathology is detected. Chronic administration of cholate increases weakly bile production in CF mice when compared with control mice. Moreover, by lowering the phospholipid-to-bile salt ratio, chronic cholate administration increases the cytotoxicity of bile produced by CF mice. By contrast, either acute or chronic ursodeoxycholic acid (UDCA) administration induces a bile salt-specific choleresis in vivo in mice that is independent of functional CFTR [44]. These observations may help to explain the beneficial effects of oral UDCA therapy in cholangiopathies including CFLD. 4.5. The gallbladder In mice, the gallbladder is an accessible and relatively simple epithelial tissue with both secretory and absorptive properties. We have previously shown that CFTR-dependent forskolin-induced Isc ( Isc forsk) is absent in Cftrtm1Cam C57Bl/6/129 knockout mice, but not normal mice [45,46]. We
  8. 8. M. Wilke et al. / Journal of Cystic Fibrosis Volume 10 Suppl 2 (2011) S152–S171 S159 Fig. 4. Isoproterenol-stimulated salivary gland secretion. Salivary gland secretion was measured in anesthetized mice for 21 minutes as described in [38]. Mice were pretreated with atropine to block cross-stimulation of cholinergic-stimulated secretion. Isoproterenol was used in the presence of atropine to stimulate adrenergic secretion. There was no significant difference between the maximum secretion of FVB wild-type (WT) male (n = 6) and female mice (n = 4). The CF mutant homozygous Cftrtm1Eur FVB F508del (DF/DF) littermates (DF/DF: male: n = 123, female: n = 63) and homozygous Cftrtm1Cam FVB knockout mice (−/−: male: n = 7, female: n = 11) of either sex had a very low response compare to wild-type (P < 0.001). The Cftrtm1Eur F508del mice do not differ from the Cftrtm1Cam. further showed that unstimulated gallbladders from normal mice reabsorb fluid to concentrate bile, whereas they secrete fluid when stimulated with forskolin [45]. By contrast, in gallbladders from Cftrtm1Cam knockout mice forskolin abol- ished fluid reabsorption without stimulating fluid secretion, demonstrating that fluid secretion is CFTR-dependent [45]. In Cftrtm1Eur 129/FVB F508del mice, forskolin-elicited Isc responses were reduced by 80% [9]. Because Cftrtm1Cam FVB mice also show low, but detectable, forskolin-elicited Isc responses through a CFTR-independent pathway, the residual forskolin-induced Isc is not mediated by F508del-CFTR (De Jonge, Bot unpublished data). Here, we present previously unpublished data demonstrating that the forskolin-induced Fig. 5. Fluid transport in mouse gallbladder. Net fluid transport rates (Jv) in isolated gallbladders of the 129/FVB Cftrtm1Eur F508del strain [10] were measured as described in [45]. Negative values show initial transport of fluid from the lumen of the bladder to the bath. After addition of the cAMP agonist forskolin, Jv changes to positive values in normal mice, i.e. net fluid secretion towards the lumen. By contrast, in homozygous Cftrtm1Eur F508del littermates transport towards the bath is inhibited and there is no net secretion into the lumen. inhibition of reabsorption and absence of fluid secretion is also observed in Cftrtm1Eur 129/FVB F508del mice (Fig. 5). This indicates that CFTR activity is insufficient for net fluid secretion in this CF mutant mouse. Gallbladder epithelial cells contain a large number of secretory granules containing mucins. Peters et al. [46] demonstrated that basal and ATP- stimulated mucus release from cultured gallbladder epithelial cells of Cftrtm1Eur 129/FVB mice are quantitatively normal. However, the potential role of CFTR-mediated bicarbonate se- cretion in mucus release in this model system and the physical properties of the secreted mucus remain to be investigated. 4.6. Smooth muscle cells and CFTR function CFTR is expressed in smooth muscle cells and skeletal muscle myotubes, colocalised with the sarcoplasmic reticu- lum [47–50]. Recently Divangahi et al. [47] demonstrated that CFTR deficient myotubes display increased Ca2+ signaling upon depolarisation and excessive NF-κB translocation in re- sponse to inflammatory stimuli. Furthermore, degeneration of diaphragmatic muscle was observed selectively in Cftrtm1Unc mice challenged with Pseudomonas aeruginosa, which chron- ically infects the lungs of CF patients [47]. The authors suggest therefore that reduced muscle strength, particularly of the diaphragm, as frequently observed in CF patients, may contribute to CF lung disease. Because smooth muscle cells of the airways and arteries express functional CFTR [48–50], elements of the CF lung phenotype (e.g. wheezing, hyperreactivity and progressive tissue remodeling) might also result from the dysfunction of smooth muscle. 4.7. The cellular immune system is abnormal in CF mutant mice Several authors have reported abnormal behaviour of alveolar and peritoneal macrophages in CF mutant mice. Di et al. [51] demonstrated defective killing of bacteria after uptake, possibly due to dysfunctional CFTR-dependent acidification of endosomes/phagosomes [52], though this explanation is contested [53,54]. Leal and co-authors showed that peritoneal and alveolar macrophages from Cftrtm1Eur mice display a pro- inflammatory bias after stimulation in culture, characterized by enhanced secretion of IL1β, but reduced secretion of IL10 [55,56]. Similarly, Bruscia et al. [57] reported abnormal pro- inflammatory cytokine secretion by lung and bone marrow- derived macrophages in CF mutant mice after exposure to lipopolysaccharide from Pseudomonas aeruginosa (PA LPS). These data strongly suggest that macrophages from CFTR mutant mice display defects that might, at least in part, explain the pro-inflammatory lung phenotype in CF mutant mice. Other cells of the myeloid lineage (e.g. neutrophils [58]; mast cells [59,60] and dendritic cells [61]) are now under investigation in CF patients and mouse models. If it can be demonstrated that such defects are prominent in CF patients and cell-autonomous, not adaptive, in origin, this would support the conclusion that CF is a disease of the immune system.
  9. 9. S160 M. Wilke et al. / Journal of Cystic Fibrosis Volume 10 Suppl 2 (2011) S152–S171 4.8. Abnormal bone and cartilage formation in CF mutant mice Bonvin et al. [62] first showed that rings of tracheal carti- lage in CF mutant mice (Cftrtm1Unc knockout and Cftrtm1Eur F508del (129/FVB)) are frequently incomplete at birth. Be- cause, neonatal CFTR knockout ferrets [6] and pigs [4,5] also exhibit abnormal tracheal cartilage, this developmental abnor- mality in CFTR deficient animals is the first to be verified in multiple species. An explanation for this phenomenon is not readily available. Interestingly, mouse chondrocytes, express functional CFTR [63], suggesting a cell autonomous defect in cartilage formation. However, a response to local and sys- temic inflammation cannot be excluded. Of note, a similar, but equally unexplained, phenotype was reported in TMEM16A mutant mice [64–66], which lack the Ca2+-activated Cl− channel (see above, section 4.12). Reduced bone density in CFTR mutant mice was described initially by Dif et al. [67] and confirmed by others [68,69]. It is currently unknown whether this defect is caused by a cell autonomous CFTR-related defect of bone forming osteoblasts or bone degrading osteoclasts. Alternatively, this defect might be secondary to systemic inflammation and abnormal bioactive lipid metabolism (ceramides, vitamin D). Low amounts of CFTR protein are expressed in mouse and human osteoblasts and odontoblasts, together with low levels of CFTR mRNA [70]. The location of CFTR in these cells was mainly intracellular, suggesting that CFTR might function as an anion transporter in intracellular organelles. The dysfunction of the ClC transporter ClC-7 in osteopetrosis [71,72] raises the possibility that CFTR dysfunction in organelle membranes of bone cells might underlie the changes in bone metabolism in CFTR-deficient mice and CF patients, leading to osteopenia. The clinical relevance of these data remains to be estab- lished. To date, no cartilage abnormalities have been reported in CF patients. However, osteopenia and osteoporosis are serious problems in CF patients [73,74]. Taken together, the data suggest that animal model studies of CF-related bone disease might lead to novel treatments for CF patients. 4.9. Nasal airways of CF mutant mice The nasal epithelium is an important focus of CF mouse model research, because of its similarities in structure and function with the human bronchial airway epithelium. Indeed, early studies of the nasal epithelium demonstrated electro- physiological abnormalities consistent with CFTR dysfunc- tion in CF mutant mice (i.e. abrogated cAMP-stimulated Cl− conductance and enhanced amiloride-sensitive Na+ currents in microperfusion studies, Table 2). More recent studies draw attention to the complex ar- chitecture of the mouse nasal epithelium and the effects of CFTR dysfunction on this tissue [75,76]. The mouse nasal cavity is a septated cartilage and bone structure covered with nasal airway epithelium (NAE), mainly ciliated and mucus producing (goblet) cells. However, the cavity lining the up- per part of the skull is covered with olfactory epithelium (OE). The OE consists of a multilayer of neuronal cells projecting sensory cilia from the surface, covered with a layer of specialized epithelial cells called sustentacular cells (Fig. 6A and B). This tissue expresses CFTR and ENaC and is thought to contribute to fluid and ion homeostasis across the OE [76]. Under the OE and the NAE, there are numerous submucosal glands embedded in connective tissue. CF mutant mice display an age-dependent progressive degeneration of the olfactory neuronal layer that is evident in adult mutant mice from the age of 16–24 weeks. This was first reported for Cftrtm1Unc knockout and CftrTgHm1 G551D mutant mice on different genetic backgrounds [75,76]. Furthermore, an increase of NAE thickness and apparent hyperplasia of mucus producing cells is observed in CftrTgHm1 G551D mutants [75]. At age 14–20 weeks, Cftrtm1Eur FVB mice did not show histological evidence of OE degeneration in our facility (Fig. 6A, B), indicating that this abnormality is not only age-, but also mouse strain-dependent. When the OE and underlying tissue is removed from the nasal cavity, it can be inserted into a small aperture Ussing chamber for ion transport studies [77]. In all five CF mutant strains tested, we observed a significantly increased (P < 0.01) Isc response to amiloride ( Isc amil) (Fig. 7). This includes the Cftrtm1Eur F508del mutation on different genetic backgrounds and the Cftrtm1Cam knockout allele on mixed background (C57Bl/6-129) and FVB. An enhanced Isc amil and PDamil is also observed in human CF airway epithelia, supporting the hypothesis that CF airway disease results, at least in part, from ENaC hyperactivity. Consequently, ENaC is a primary therapeutic target in CF mutant mice [15,78]. In excised nasal OE from Cftrtm1Eur F508del mice, Cftr mRNA expression is normal (Scholte et al., data not shown), whereas CFTR protein expression is much reduced (Fig. 6A and B) and in Cftrtm1Cam knockout mice this signal is absent (not shown). The absence of olfactory marker proteins from CFTR-expressing cells argues that CFTR is not expressed in ciliated sensory neurons, but in a subset of microvillous secondary chemosensory cells (named brush cells), where it is co-localized with ENaC (De Jonge et al, unpublished data). Whether this cell type represents a cell specialized for fluid and electrolyte transport or plays a role in CFTR-dependent regeneration of the olfactory system [79,80] remains to be established. Forskolin responses observed in different mouse strains are complex parameters that cannot, at face value, be attributed to CFTR activity. In Cftrtm1Cam C57Bl/6-129 knockout mice, we observe a forskolin response comparable to normal littermates. However, in Cftrtm1Cam FVB mutant mice this response is virtually absent (Fig. 7). These data suggest that a cAMP-stimulated Cl− conductance other than CFTR is present in the OE of Cftrtm1Cam C57Bl/6-129 mice, but is absent from Cftrtm1Cam FVB mice. Studies are underway to identify this cAMP-stimulated Cl− conductance. In all Cftrtm1Eur F508del backcrossed strains the nasal forskolin-stimulated Isc in mutant mice is comparable to normal littermates. Because this includes the FVB strain
  10. 10. M. Wilke et al. / Journal of Cystic Fibrosis Volume 10 Suppl 2 (2011) S152–S171 S161 Fig. 6. CFTR immunostaining in mouse olfactory epithelium (OE) and trachea. In wild-type FVB (FVB WT) mice punctuated CFTR staining in brush cells of the olfactory epithelium (OE) (A) and diffuse CFTR staining in ciliated cells of the trachea (C) is observed. In homozygous mutant littermates (FVB F508del Cftrtm1Eur), the CFTR staining in brush cells of the OE (B) and ciliated cells of the trachea (D) and is much reduced. The CFTR signal is completely absent in sections from Cftrtm1Cam FVB knockout mice, stained in parallel (not shown). that does not show a response with the Cftrtm1Cam knockout allele, one possible explanation is that the nasal forskolin- stimulated Isc reflects residual CFTR activity resulting from partial processing of the mutant protein [9]. However, the nasal forskolin-stimulated Isc does not influence the enhanced amiloride response in this mutant strain (Fig. 7). Together with the immunohistology data (Fig. 6A, B), this argues that there is not a linear relationship between the nasal forskolin- stimulated Isc and the number of active CFTR channels. 4.10. Submucosal glands in CF mutant mice Mice have composite serous and mucous submucosal glands, not only in the nasal cavity, but also in the upper part of the trachea. This is relevant, because abnormal submucosal glands are thought to play an important role in the development of CF airway disease, contributing to fluid hypo-secretion and the abnormal consistency of mucus in the proximal airways [81,82]. Secretion of fluid from submucosal glands is dependent on two complementary combinations of transport systems, in particular basolateral K+ channels and apical CFTR Cl− channels, regulated by the intracellular calcium and cAMP signaling pathways, respectively. The cAMP-stimulated component is lacking in tracheal glands from CF mutant mice, and is therefore thought to be CFTR-dependent [81,82]. The contribution of CFTR-mediated secretion by submucosal glands to CF mouse pathology is unclear, but this experimental paradigm constitutes a valid model to study CFTR function and test drugs targeting CFTR. It is important to note that the submucosal glands of the nasal cavity, and their contribution to its CF phenotype, have not yet been studied. It is conceivable, that nasal submucosal cells have properties different from tracheal glands, since they have developed in a different cellular context. 4.11. CFTR expression in the mouse trachea The cellular architecture of the pseudostratified columnar mouse tracheal epithelium is similar to that of the nasal airways. It is dominated by CFTR-expressing ciliated cells (Fig. 6C and D), interspersed with infrequent mucus secreting neuroendocrine and Clara cells. Mucociliary clearance and epithelial ion transport are not significantly affected in tracheal
  11. 11. S162 M. Wilke et al. / Journal of Cystic Fibrosis Volume 10 Suppl 2 (2011) S152–S171 Fig. 7. Summary of short-circuit current (Isc) measurements in nasal epithelium. Short-circuit current responses ( Isc) to amiloride (10 μM) and forskolin (10 μM) were measured in isolated nasal epithelium in an Ussing chamber as described in [77]. Bioelectric characteristics of 12–18 weeks old normal (nn) and homozygous mutant (dd) mice from the different backcrossed Cftrtm1Eur F508del strains (FVB, FVB/129, C57Bl/6, 129sv) are compared. Homozygous Cftrtm1Cam knockout mice were analysed in FVB (FVB/n -/-), and the original Cftrtm1Cam strain (C57Bl/6-129 -/-), the latter compared to their normal (nn) littermates. The number of individual animals measured per experimental group is given in parenthesis. The amiloride response is significantly (P < 0.01) increased in all Cftr mutant strains compared to their wild-type littermates, except for the 129sv backcross. The forskolin response does not differ significantly between the mutant strains and the wild-type mice in this assay, with the exception of the FVB Cftrtm1Cam knockout mouse strain (P < 0.001). epithelia from CFTR knockout mice [83], but see [84]. Interestingly, in tracheal epithelia the Ca2+-activated Cl− channel (CaCC) is the dominant apical membrane Cl− channel, CFTR activity is only observed after maximal stimulation of CaCC [85]. 4.12. CaCC: the dark horse in CF pathology CaCC is expressed in the apical membrane of epithelia affected by CF, where it operates in parallel with CFTR, but is regulated by a different intracellular signaling pathway. As a result, CaCC is a potentially important element in CF pathology. The recent demonstration that the TMEM16A gene encodes a CaCC [86–88] will allow the molecular behaviour of CaCC in different secretory epithelia to be elucidated. Interestingly, TMEM16A knockout mice exhibit phenotypic abnormalities consistent with a role in fluid and electrolyte homeostasis [64–66]. It is therefore likely that the phenotype of both CF mutant mice and CF patients is, at least in part, dependent on TMEM16A expression and function. To what extent CaCC affects the phenotypic abnormalities observed in different organs of CF mice, as well as in CF patients, remains to be established. 4.13. CFTR expression in the mouse lung In mouse lung, Cftr mRNA expression is readily detected by RT-PCR and in situ hybridization [89], but expression levels are 10-fold lower than in nasal epithelium (Scholte et al., not shown). Immunohistological detection of CFTR ex- pression in the lung, confirmed by using CFTR knockout mice as a negative control, is not feasible with current methods. Mouse bronchi possess ciliated and mucus producing cells in a declining proximal to distal distribution, whereas airways in the mouse lung are dominated by Clara cells, constituting 60–80% of the total epithelial cell population. Clara cells are non-ciliated secretory cells that (i) display phenotypic adapta- tion and proliferative capacity during injury and inflammation, (ii) are capable of adopting a mucus secretory phenotype [90– 92] and (iii) express functional CFTR [93,94]. Clara cells are also prominent in human bronchioles, which, small in size, but large in number, comprise a major portion of the airway surface in the human lung. Like airways in the mouse lung, human bronchioles lack cartilage and submucosal glands. Of note, bronchioles play an important role in the initiation of CF lung disease. For example, Tiddens et al. [95] and Stick et al. [96] observed severe distal airway disease in at least 50% of CF patients before the age of 5 years, including patients free of bacterial colonization. Notwithstanding impor- tant functional differences between human and murine airway epithelia, murine nasal airway epithelia is therefore struc- turally and functionally related to human bronchial airways, whereas murine bronchi are similar to human bronchioles. In alveolar epithelia, CFTR is expressed in type II epithelial cells [97], where it is involved in cAMP-stimulated fluid reabsorption [98]. Using an ingenious optical method, Lindert et al. [99] demonstrated that basal fluid secretion in the alveolar space was absent in Cftrtm1Unc homozygous mutant
  12. 12. M. Wilke et al. / Journal of Cystic Fibrosis Volume 10 Suppl 2 (2011) S152–S171 S163 mouse lung and sensitive to the CFTR inhibitor CFTRinh-172 in normal mice. 5. Lung pathology in CF mutant mice CF lung pathology has been studied intensively using CF mutant mice and has been thoroughly reviewed [1–3]. Here we present a summary of the literature and some relevant recent data. Two issues are prominent in this field. First, to what extent can we compare lung pathology between humans and mice? Second, which parameters in a mouse model provide relevant information about CF pathology? In view of the major differences in size, cellular archi- tecture, physiology, host-pathogen interactions, lifestyle and lifespan, it is not surprising that CF mice do not reproduce exactly the lung disease typical of CF patients. However, there is considerable heterogeneity in the severity and progression of CF lung disease in humans. Another caveat is that we know surprisingly little about the early development of CF lung disease from direct observations of human airways in situ, especially the distal airways [95]. It is important to take these considerations into account when comparing lung pathology in mouse models and CF patients. Several groups reported an inflammatory lung phenotype associated with progressive tissue remodeling in different unchallenged CF mutant mice [55,100–103]. The severity and presentation of this phenotype, most prominent in a C57Bl/6 backcross of the Cftrtm1Unc mouse [19], appears to be strain- and age-dependent. As discussed above, this phenotype will be influenced by animal husbandry. Here, we show that the inflammatory score based on quan- titative histology is moderately, but significantly increased in adult (16–19 weeks) unchallenged Cftrtm1Eur F508del (FVB) mice kept in ventilated cages under pathogen-free conditions (Fig. 8). Inflammation, enhanced levels of the macrophage at- tractant CCL2/MCP-1 and increased numbers of macrophages were reported in lung lavage fluid of a different colony of Cftrtm1Eur F508del (129/FVB) mice, which were reduced by the macrolide azythromycin [55]. Importantly, this in- flammatory phenotype is not related to detectable pathogen colonization, suggesting that CFTR deficiency by itself is sufficient to generate a pro-inflammatory environment in the mouse lung. At present it has not been established whether this inflammatory lung phenotype is exclusively attributed to the abnormal behaviour of cells from the myeloid lineage, in particular macrophages [57,104], or whether abnormal proin- flammatory signaling in CFTR-deficient epithelial cells plays a role [104]. However, it should be noted that these animals are not kept in a sterile environment. Exposure to fecal bacteria, dust and other allergens are expected to be more challenging to CF mutant animals than to normal littermates. Therefore, it might be misleading to call the inflammatory lung phenotype “spontaneous”. Moreover, animal husbandry, genetic background and gender likely affect the data, which explains why reports from different laboratories differ with respect to the severity and presentation of the inflammatory lung phenotype. Fig. 8. Lung inflammatory score is increased in Cftrtm1Eur FVB mutant mice. Paraffin sections were made from paraformaldehyde inflated lungs of FVB Cftrtm1Eur F508del mice, homozygous normal (+/+, n = 5) and mutant (dd, n = 6), age-matched (16–19 weeks comparable numbers of males and females) littermates. The animals were kept in the Erasmus MC animal facility, pathogen-free (FELASA+), in sterilized ventilated cages, provided ad lib with normal chow and sterilized acidified water. Five micron sections were stained with hematoxylin/eosin and scored blind by an independent pathologist, for oedema, haemorrhages, recruitment of poly-morphonuclear cells, macrophages, lymphocytes and plasma cells, lesions of alveolitis and bronchitis, area of focal or diffuse consolidation, necrosis and metaplasia of pneumonocytes. The scores in each category are expressed and averaged on a scale of zero (no abnormalities) through one (mild inflammation) to five (severe, destruction of the lung with lesions of alveolitis, severe bronchitis, with necrosis and large areas of consolidation extended to all lobules). The results show mild, but significant inflammation in CF mutant mice compared to normal, under these conditions. 5.1. A challenged mouse counts for two Waiting for CF mice to develop lung disease under stan- dard laboratory conditions is a tedious, ineffective method. Therefore, CF mutant mice of various strains have been subjected to bacterial challenge protocols, resulting in a substantial, albeit bewildering, body of data. CF mice were infected with Pseudomonas aeruginosa (PA), because this opportunistic pathogen causes life threaten- ing chronic infections in CF patients. At first sight this seems a rather naïve approach, since host pathogen interactions are very complex, differ in mice and humans and are certainly not dependent on CFTR activity alone. Moreover, PA adapts rapidly to different environments, from hospital taps and shower heads to patient lungs, forming mucoid variants and biofilms. Nevertheless, several authors report enhanced mor- tality and reduced clearance of acute PA infection in loss of function CF mutants [105–108] and G551D mice [109]. What causes mortality in acute PA infection models is not obvious, and should probably be distinguished from the intensity and resolution of inflammation. Interestingly, epithelial, but not macrophage, specific Cftr gene complementation is necessary and sufficient to protect G551D CF mice from PA mortality [110]. While on the whole data from acute PA infection experi- ments are satisfactory, several concerns need to be addressed. It proved difficult to establish a true chronic infection with clinical PA isolates. Mice, including CF mice, either rapidly clear an intranasal or intra-tracheal dose of PA or they die
  13. 13. S164 M. Wilke et al. / Journal of Cystic Fibrosis Volume 10 Suppl 2 (2011) S152–S171 quickly depending on the dose and method of delivery. Dif- ferent clinical isolates and delivery methods have therefore been tested [111,112]. Virulent PA strains embedded in agar beads elicit a prolonged inflammatory reaction, simulating chronic infection. Interestingly, responses to this challenge are more severe in CF mutant mice [113], including the gut- corrected version of the Cftrtm1Unc knockout mouse [114]. In the latter model, azithromycin reduced inflammation and improved bacterial clearance after a challenge with alginate embedded PA [115]. Though this is an elegant way to create a prolonged inflammatory challenge, it does not necessarily reflect accurately the response of human CF epithelia to PA infection or how biofilms are formed and maintained in CF lungs. In a different approach, Pier and coauthors showed that delivery of PA in drinking water results in low level chronic colonization of CF mouse airways, but not those of controls [116,117]. In this model, the inability to clear infection was linked to abnormal IL-1β signaling in epithelial cells [118]. However, this model lacked the intense inflammatory responses of mice infected with high doses of PA. Intra-tracheal lipopolysaccharide (LPS) isolated from bac- teria can also be used to induce transient lung inflammation. Similar to bacterial infection, this response is characterized by activation of alveolar macrophages, cytokine production, a massive neutrophil influx, but little or no mortality. The responses seen in these models are primarily triggered by Toll like receptors (TLR) of the innate immune system. In addition to TLR2 and TLR3 responses to LPS, bacterial flagellin is recognized by TLR5 and can be used to provoke an inflam- matory response [119,120]. In CF mutant mice responses to LPS challenge are stronger than in normal mice [56,57]. 5.2. Mucus production in CF mutant mouse lung Mucus plugging of the airways is an important aspect of CF lung disease. This has been attributed to abnormal hydration of CF mucus by the surface epithelium hyperabsorbing Na+ [78] and alternatively to a lack of bicarbonate secreting cells, which prevents normal mucus hydration and secretion [33,34,121]. Furthermore, goblet cell hyperplasia and mucus hyper-secretion induced by inflammation are likely to play a role. In the CF mouse nasal cavity, goblet cell hyperplasia does not correlate with infection [75]. In distal airways of the normal mouse lung, we observe no goblet cells using PAS/Alcian Blue, Muc5AC, and Clca3/Gob5 stains. However, in proximal airways of the lung, second and third bifurcation bronchi, we do observe areas containing PAS/Alcian Blue and Clca3/Gob-5 positive cells. Kent et al. [101] observed increased PAS staining together with other lung alterations in unchallenged congenic C57Bl/6 Cftrm1Unc mice. In an independent study, Durie et al. [19] confirmed progressive lung disease and airway accumula- tion of mucus in the same mutant strain. Cftrtm1Hgu mice demonstrate mucus accumulation compared to controls after challenge with bacteria [107]. By contrast, Cressman et al. [122] reported that Cftrtm1Unc mice with different genetic backgrounds did not exhibit goblet cell hyperplasia com- pared to normal controls, and both groups of mice responded similarly to ovalbumin challenge. More recently, Touqui and colleagues observed increased numbers of mucus producing cells and MUC5AC antigen in lungs of homozygous mu- tant C57Bl/6 Cftrtm1Unc mice [123]. Of note, the authors demonstrated that this response was mediated by abnormal activation of cytosolic phospholipase A2 and could be reduced by a specific inhibitor of this enzyme [123]. Similar results were obtained in Cftrtm1Eur mice (Scholte et al., unpublished results). In summary, unchallenged CF mutant mice present a mod- est, but significant, increase in the number of mucus producing cells in the proximal airways and mucus hypersecretion in lavage. However, overt airway plugging and air trapping such as that reported in CF patients and recently in CF ferrets and pigs is not observed. Because cells that stain with standard goblet cell markers are confined to the proximal airways in the mouse lung, careful analysis of sections is required. In view of the known susceptibility of this parameter to environmental and genetic factors care must be taken to provide properly controlled conditions. The molecular biology of this response remains to be investigated. On the basis of available data it is most likely an adaptive response of Clara cells to the pro-inflammatory environment created by CFTR deficiency. Whether a similar response occurs in human distal airways is not known. Finally, newborn mice overexpressing the β subunit of ENaC exhibit a considerably stronger version of the mucus hypersecretion phenotype (for further information see the accompanying article [17]). 6. Bioactive lipids and CF lung disease CFTR mutant mice display distinct abnormalities in the metabolism of several lipids that are known to play a role in intra- and extracellular signaling of inflammation and apoptosis. Freedman et al. [124,125] first reported that Cftrtm1Unc C57Bl/6 knockout mice show an increased ratio of arachidonic acid (AA) to docosahexaenoic acid (DHA), which was reversed by oral DHA. Importantly, this finding correlated with a normalization of several phenotypic param- eters in CF mutant mice [126]. More recently, Radzioch and colleagues confirmed the abnormal AA/DHA ratio in adult Cftrtm1Unc C57Bl/6 CF mutant mice [101] and CF patients [127]. Furthermore, these authors reported reduced levels of major ceramide species in plasma and tissue samples in both species, which were reversed by treatment with the vita- min A analogue fenretinide (N-(4-hydroxyphenyl)retinamide; 4-HPR) [127]. In CF mutant mice, fenretinide reduced the sensitivity to intranasal challenge with PA [103] and reduced bone abnormalities in CF knockout mice [128]. Whether fenretinide can reduce the development of progressive CF lung disease in patients remains to be established. Also the molecular mechanism explaining the relationship between this aspect of the CF phenotype and CFTR deficiency remains
  14. 14. M. Wilke et al. / Journal of Cystic Fibrosis Volume 10 Suppl 2 (2011) S152–S171 S165 to be determined. It is hypothesized that both oral DHA and fenretinide act by reducing pro-inflammatory signaling through the PPAR pathway. In this context, Harmon et al. [129] recently demonstrated that Cftrtm1Unc (B6.129P2) ho- mozygous knockout mice (males and females, four weeks old) have attenuated PPARγ activity due to reduced levels of PPAR agonists, including 15-keto-PGE2. The authors further showed that the PPARγ agonist rosiglitazone, used to treat diabetes, corrected inflammatory and intestinal pathology in the mouse model [129]. In accordance with these observa- tions, we have reported previously that expression of the 15-keto-PGE2 precursor, PGF2α and Cbr2 (lung carbonyl reductase), the enzyme responsible for PGF2α production in airway epithelial cells, are reduced in BALF of homozygous Cftrtm1Eur lungs [130]. How these findings relate to defective CFTR function is not yet fully understood. Borot et al. [131] presented evidence suggesting that CFTR can form a com- plex with cytosolic phospholipase A2 through the S100A10 adaptor protein, which might explain the hyperactivity of cellular phospholipase A2 in CF mutant mice and its effect on mucin secretion [123]. We hypothesize that CFTR dysfunc- tion affects the enzymatic pathways involved, either directly or indirectly, through an adaptive process. In a different CF mouse model carrying the Cftrtm1Hgu partially active knockout allele (Table 1), Gulbins and col- leagues reported an age-dependent accumulation of ceramide species in lung tissue, which correlates with progressive lung inflammation and sensitivity to PA challenge [102]. Reduction of acid sphingomyelinase activity (ASM), either by mutation (Smpd−/−) or using the ASM inhibitor amitryptilin normal- ized ceramide levels and reduced lung pathology in these CF mutant mice [102]. Further studies confirmed and extended these data in the mouse model [132]; initial experiments in CF patients demonstrated that amitriptylin is safe and might be effective in reducing CF disease [133]. At this time, the apparent contradictory changes in ceramide levels in CF mutant mice between the Radzioch and Gulbins groups have not been resolved using verified experimental protocols. It is possible that differences in experimental protocols may cause a bias towards distinct subsets of ceramide species. 7. Pumping the pipeline, testing experimental therapies in mouse models A key application of CF mutant mice is to test experimental therapeutics. Elsewhere in this issue the development of pharmacotherapy [134] and gene therapy for CF [135] are reviewed. New antagonists of ENaC have been developed and are being tested in a mouse model of ENaC hyperactivity [17]. Compounds affecting bioactive lipid metabolism and downstream transcription factor activity were discussed in the previous section. The cloning of the CaCC TMEM16A also presents new opportunities for intervention. As discussed below, several experimental therapeutics have been tested in CF mouse models. 7.1. CFTR activators The mutant strains bearing F508del and G551D were made specifically to test compounds that aim to correct defec- tive processing and trafficking (CFTR correctors) or defective channel gating (CFTR potentiators). A caveat is that these mu- tations are introduced into the mouse CFTR sequence, which differs from the human sequence. Therefore, the properties of the mutant protein, including its interaction with therapeutic drugs, might differ from the human form. Indeed, the murine CFTR Cl− channel exhibits significant differences in its bio- physical properties and regulation compared to human CFTR. These differences include (i) a smaller single-channel conduc- tance, (ii) a markedly different pattern of channel gating and (iii) insensitivity to AMP-PNP and pyrophosphate (PPi), two agents that potentiate robustly the activity of human CFTR [136–138]. Consistent with these data, a number of CFTR potentiators (e.g. VRT-532, NPPB-AM and VX-770), which augment strongly human CFTR channel gating are without effect on murine CFTR [139,140]. Moreover, Scott-Ward et al. [141] demonstrated that transfer of both nucleotide-binding domains (NBDs) of murine CFTR to human CFTR endowed the human CFTR Cl− channel with the gating behaviour and pharmacological properties of the murine CFTR Cl− chan- nel. These data suggest that human-murine CFTR chimeras might be employed to identify the binding sites of CFTR potentiators. 7.2. CFTR correctors Because CFTR correctors target the cellular protein pro- cessing and trafficking machinery, they are less dependent on a direct interaction with the CFTR protein. As a re- sult, a number of CFTR correctors have been successfully tested in the F508del mouse model (De Jonge et al., in preparation). For example, miglustat, currently under eval- uation as a therapy for Gauchers disease, shows limited, but significant, efficacy in the Cftrtm1Eur mouse model by reducing the amiloride response in the nasal epithelium [142] and increasing F508del CFTR processing efficiency in intestinal epithelium [143]. Moreover, intraperitoneal instil- lation of sildenafil and vardenafil, related phosphodiesterase 5 inhibitors, activated nasal Cl− conductance in Cftrtm1Eur F508del mice, but not Cftrtm1Cam null mice with the same genetic background [144]. In a recent study, Robert et al. [145] showed that glafenine, an off-patent drug with analgesic properties, improved the trafficking of human F508del-CFTR in BHK and CFBE41o− cells to 40% of normal values. In isolated intestinal epithelia from Cftrtm1Eur mice, glafe- nine increased the Isc response to forskolin plus genistein, and it partially restored salivary secretion in vivo in mutant mice, without causing adverse effects [145]. These results illustrate that mutant CFTR mouse strains can play a signif- icant role in testing experimental drugs that are identified in high-throughput screening programs [134].
  15. 15. S166 M. Wilke et al. / Journal of Cystic Fibrosis Volume 10 Suppl 2 (2011) S152–S171 7.3. Anti-inflammatory drugs In addition to drugs targeting the bioactive lipid pathways (see above, section 6), anti-inflammatory drugs can be investi- gated successfully in mouse models. The macrolide antibiotic azithromycin has anti-inflammatory action and reduced the hyper inflammatory condition in Cftrtm1Eur F508del mutant mice [55], apparently through an action at the level of alve- olar macrophages [146], rather than at the level of epithelial cells [104]. Similar results were reported for the Cftrtm1Unc- TgN(FABPCFTR) gut corrected CFTR null mouse model challenged with a clinical PA isolate [115]. The synthetic triterpenoid CDDO reduced the hyper inflammatory pheno- type of a R117H mutant mouse model, presumably through inhibition of the NF-κB pro-inflammatory pathway [147]. 8. Modifier genes and new therapeutic targets Not only environmental factors but also genetic elements determine CF pathology. First, the many different mutations in the CFTR gene found in CF patients give rise to variable levels of residual activity [27,148]. A different category of genetic variations, often referred to as “modifier genes”, are polymorphisms in genes that affect CFTR function or pathways involved in CF pathology [149–151]. Identification of modifier genes not only improves the CF diagnosis, but also identifies novel therapeutic targets. CF mouse models and the toolbox allowing genetic experiments in this species offer unique opportunities to study modifier genes and their potential therapeutic effects. First, phenotypic differences between strains in different genetic backgrounds can be analyzed [152,153] followed by gene mapping [154–156]. Current developments in gene and transcriptome sequencing technology (“deep sequencing”) are likely to boost this field. Another approach is to create specific mutations in candidate genes that emerge from genetic or physiological studies in the CF patient population. As an example, the identification of the neutrophil marker IFRD1 as a modifier of lung disease in CF patients was supported by analysis of Ifrd1 mutants in mice [58]. Moreover, using KCNE3 knockout mice, Preston et al. [157] recently demonstrated that the K+ channel β subunit KCNE3 might act as a modifier gene in CF. Heteromultimeric KCNQ1/KCNE3 K+ channels provide the driving force for Cl− exit across the apical membrane of airway and intestinal epithelia. In KCNE3 knockout mice, CFTR-mediated Cl− secretion is attenuated markedly [157]. The availability of an increasing number of conditional and inducible mutant mouse strains, allows further validation of putative target genes. The EUCOMM program (, which aims to generate conditional mutant embryo stem cell lines of all functional genes in the mouse genome is linked to the EUMODIC program ( which coordinates phenotypic analysis of derived strains. Crossbreeding of selected mutations with mutant Cftr alleles is expected to improve our knowledge of CF pathology. 9. Conclusion and perspective We can conclude that despite their obvious limitations, Cftr mutant mouse models have delivered. First, a better understanding of CF pathology beyond clinical and cell culture studies. Second, novel and promising therapeutics have been identified, which are now being tested in these models. Large animal models will certainly offer exciting new dimensions in this field. However, the versatility and relative affordability of mouse strains in genetic and physiological studies will ensure their continued use. Acknowledgements We thank EuroCareCF (LSHM-CT-2005-018932), the Dutch CF Foundation NCFS (R.B-O.) and the Dutch Maag Lever Darm Stichting MLDS (M.W.), for financial support. We gratefully acknowledge the expert assistance of Michel Huerre (Institut Pasteur, Paris, France) in the quantitative analysis of histological data. Conflict of interest All authors report no conflict of interest References [1] Grubb BR, Boucher RC. Pathophysiology of gene-targeted mouse models for cystic fibrosis. Physiol Rev 1999;79(1 Suppl):S193–214. [2] Scholte BJ, Davidson DJ, Wilke M, De Jonge HR. Animal models of cystic fibrosis. J Cyst Fibros 2004;3(Suppl 2):183–90. [3] Guilbault C, Saeed Z, Downey GP, Radzioch D. Cystic fibrosis mouse models. Am J Respir Cell Mol Biol 2007;36:1–7. [4] Rogers CS, Hao Y, Rokhlina T, et al. Production of CFTR-null and CFTR- F508 heterozygous pigs by adeno-associated virus- mediated gene targeting and somatic cell nuclear transfer. J Clin Invest 2008;118:1571–7. [5] Rogers CS, Abraham WM, Brogden KA, et al. The porcine lung as a potential model for cystic fibrosis. Am J Physiol Lung Cell Mol Physiol 2008;295:L240–63. [6] Sun X, Yan Z, Yi Y, et al. Adeno-associated virus-targeted disruption of the CFTR gene in cloned ferrets. J Clin Invest 2008;118:1578–83. [7] Meyerholz DK, Stoltz DA, Pezzulo AA, Welsh MJ. Pathology of gastrointestinal organs in a porcine model of cystic fibrosis. Am J Pathol 2010;176:1377–89. [8] Stoltz DA, Meyerholz DK, Pezzulo AA, et al. Cystic fibrosis pigs develop lung disease and exhibit defective bacterial eradication at birth. Sci Transl Med 2010;2:29ra31. [9] French PJ, van Doorninck JH, Peters RH, et al. A F508 mutation in mouse cystic fibrosis transmembrane conductance regulator results in a temperature-sensitive processing defect in vivo. J Clin Invest 1996;98:1304–12. [10] van Doorninck JH, French PJ, Verbeek E, et al. A mouse model for the cystic fibrosis F508 mutation. EMBO J 1995;14:4403–11. [11] Delaney SJ, Alton EW, Smith SN, et al. Cystic fibrosis mice carrying the missense mutation G551D replicate human genotype-phenotype correlations. EMBO J 1996;15:955–63. [12] Hodges CA, Cotton CU, Palmert MR, Drumm ML. Generation of a conditional null allele for Cftr in mice. Genesis 2008;46:546–52. [13] Mall M, Grubb BR, Harkema JR, O’Neal WK, Boucher RC. Increased airway epithelial Na+ absorption produces cystic fibrosis-like lung disease in mice. Nat Med 2004;10:487–93. [14] Mall MA, Harkema JR, Trojanek JB, et al. Development of chronic
  16. 16. M. Wilke et al. / Journal of Cystic Fibrosis Volume 10 Suppl 2 (2011) S152–S171 S167 bronchitis and emphysema in β-epithelial Na+ channel-overexpressing mice. Am J Respir Crit Care Med 2008;177:730–42. [15] Mall MA. Role of the amiloride-sensitive epithelial Na+ channel in the pathogenesis and as a therapeutic target for cystic fibrosis lung disease. Exp Physiol 2009;94:171–4. [16] Zhou Z, Treis D, Schubert SC, et al. Preventive but not late amiloride therapy reduces morbidity and mortality of lung disease in βENaC- overexpressing mice. Am J Respir Crit Care Med 2008;178:1245–56. [17] Zhou Z, Duerr J, Johannesson B, et al. The βENaC-overexpressing mouse as a model of cystic fibrosis lung disease. J Cyst Fibros 2011;10(Suppl 2):172–82. [18] Reynaert I, Van Der Schueren B, Degeest G, et al. Morphological changes in the vas deferens and expression of the cystic fibro- sis transmembrane conductance regulator (CFTR) in control, F508 and knock-out CFTR mice during postnatal life. Mol Reprod Dev 2000;55:125–35. [19] Durie PR, Kent G, Phillips MJ, Ackerley CA. Characteristic mul- tiorgan pathology of cystic fibrosis in a long-living cystic fibro- sis transmembrane regulator knockout murine model. Am J Pathol 2004;164:1481–93. [20] Wright JT, Kiefer CL, Hall KI, Grubb BR. Abnormal enamel de- velopment in a cystic fibrosis transgenic mouse model. J Dent Res 1996;75:966–73. [21] Chang HY, Mitzner W. Sex differences in mouse models of asthma. Can J Physiol Pharmacol 2007;85:1226–35. [22] Kent G, Oliver M, Foskett JK, et al. Phenotypic abnormalities in long-term surviving cystic fibrosis mice. Pediatr Res 1996;40:233–41. [23] Cottart CH, Bonvin E, Rey C, et al. Impact of nutrition on phenotype in CFTR-deficient mice. Pediatr Res 2007;62:528–32. [24] Werner A, Bongers ME, Bijvelds MJ, de Jonge HR, Verkade HJ. No indications for altered essential fatty acid metabolism in two murine models for cystic fibrosis. J Lipid Res 2004;45:2277–86. [25] Bijvelds MJ, Bronsveld I, Havinga R, Sinaasappel M, de Jonge HR, Verkade HJ. Fat absorption in cystic fibrosis mice is impeded by defective lipolysis and post-lipolytic events. Am J Physiol Gastrointest Liver Physiol 2005;288:G646–53. [26] Ohlsson L, Hjelte L, Huhn M, et al. Expression of intestinal and lung alkaline sphingomyelinase and neutral ceramidase in cystic fibrosis F508del transgenic mice. J Pediatr Gastroenterol Nutr 2008;47:547– 54. [27] Rowntree RK, Harris A. The phenotypic consequences of CFTR mutations. Ann Hum Genet 2003;67:471–85. [28] Thiagarajah JR, Verkman AS. CFTR pharmacology and its role in intestinal fluid secretion. Curr Opin Pharmacol 2003;3:594–9. [29] Valverde MA, Vazquez E, Munoz FJ, et al. Murine CFTR channel and its role in regulatory volume decrease of small intestine crypts. Cell Physiol Biochem 2000;10:321–8. [30] Venkatasubramanian J, Ao M, Rao MC. Ion transport in the small intestine. Curr Opin Gastroenterol 2010;26:123–8. [31] Kunzelmann K, Mall M. Electrolyte transport in the mammalian colon: mechanisms and implications for disease. Physiol Rev 2002;82:245–89. [32] Leverkoehne I, Holle H, Anton F, Gruber AD. Differential expression of calcium-activated chloride channels (CLCA) gene family members in the small intestine of cystic fibrosis mouse models. Histochem Cell Biol 2006;126:239–50 [33] Quinton PM. Birth of Mucus. Am J Physiol Lung Cell Mol Physiol 2009;298:L13-4. [34] Garcia MA, Yang N, Quinton PM. Normal mouse intestinal mucus release requires cystic fibrosis transmembrane regulator-dependent bicarbonate secretion. J Clin Invest 2009;119:2613–22. [35] Malmberg EK, Noaksson KA, Phillipson M, et al. Increased levels of mucins in the cystic fibrosis mouse small intestine and modulator effects of the Muc1 mucin expression. Am J Physiol Gastrointest Liver Physiol 2006;291:G203–10. [36] Charizopoulou N, Wilke M, Dorsch M, et al. Spontaneous rescue from cystic fibrosis in a mouse model. BMC Genet 2006;7:18. [37] Wilke M, Bot A, Jorna H, Rensen S, De Jonge H. Correction of the dF508-CFTR processing defect by low temperature incubation of mouse intestine. Pediatr Pulmonol Suppl 2003;36:196 [38] Best JA, Quinton PM. Salivary secretion assay for drug efficacy for cystic fibrosis in mice. Exp Physiol 2005;90:189–93. [39] Noël S, Strale PO, Dannhoffer L, et al. Stimulation of salivary secretion in vivo by CFTR potentiators in Cftr+/+ and Cftr−/− mice. J Cyst Fibros 2008;7:128–33. [40] Catalán MA, Nakamoto T, Gonzalez-Begne M, et al. Cftr and ENaC ion channels mediate NaCl absorption in the mouse submandibular gland. J Physiol 2010;588:713–24. [41] Jouret F, Bernard A, Hermans C, et al. Cystic fibrosis is associated with a defect in apical receptor-mediated endocytosis in mouse and human kidney. J Am Soc Nephrol 2007;18:707–18. [42] Jouret F, Devuyst O. CFTR and defective endocytosis: new insights in the renal phenotype of cystic fibrosis. Pflugers Arch 2009;457:1227– 36. [43] Feranchak AP. Hepatobiliary complications of cystic fibrosis. Curr Gastroenterol Rep 2004;6:231–9. [44] Bodewes F, Wouthuizen Bakker M, Havinga R, Bijvelds M, de Jonge H, Verkade HJ. Treatment with ursodeoxycholate, but not with cholic acid, increases bile flow independently of cystic fibrosis trans- membrane conductance regulator in mice. Pediatr Pulmonol Suppl 2007;30:376. [45] Peters RH, van Doorninck JH, French PJ, et al. Cystic fibrosis transmembrane conductance regulator mediates the cyclic adenosine monophosphate-induced fluid secretion but not the inhibition of re- sorption in mouse gallbladder epithelium. Hepatology 1997;25:270–7. [46] Peters RH, French PJ, van Doorninck JH, et al. CFTR expression and mucin secretion in cultured mouse gallbladder epithelial cells. Am J Physiol Gastrointest Liver Physiol 1996;271:G1074–83. [47] Divangahi M, Balghi H, Danialou G, et al. Lack of CFTR in skeletal muscle predisposes to muscle wasting and diaphragm muscle pump failure in cystic fibrosis mice. PLoS Genet 2009;5:e1000586. [48] Vandebrouck C, Melin P, Norez C, et al. Evidence that CFTR is expressed in rat tracheal smooth muscle cells and contributes to bronchodilation. Respir Res 2006;7:113. [49] Robert R, Savineau JP, Norez C, Becq F, Guibert C. Expression and function of cystic fibrosis transmembrane conductance regulator in rat intrapulmonary arteries. Eur Respir J 2007;30:857–64. [50] Michoud MC, Robert R, Hassan M, et al. Role of the cystic fibrosis transmembrane conductance channel in human airway smooth muscle. Am J Respir Cell Mol Biol 2009;40:217–22. [51] Di A, Brown ME, Deriy LV, et al. CFTR regulates phagosome acidification in macrophages and alters bactericidal activity. Nat Cell Biol 2006;8:933–44. [52] Deriy LV, Gomez EA, Zhang G, et al. Diseasecausing mutations in the cystic fibrosis transmembrane conductance regulator deter- mine the functional responses of alveolar macrophages. J Biol Chem 2009;284:35926–38. [53] Haggie PM, Verkman AS. Defective organellar acidification as a cause of cystic fibrosis lung disease: reexamination of a recurring hypothesis. Am J Physiol Lung Cell Mol Physiol 2009;296:L859–67. [54] Barriere H, Bagdany M, Bossard F, et al. Revisiting the role of cystic fibrosis transmembrane conductance regulator and counterion permeability in the pH regulation of endocytic organelles. Mol Biol Cell 2009;20:3125–41. [55] Legssyer R, Huaux F, Lebacq J, et al. Azithromycin reduces sponta- neous and induced inflammation in F508 cystic fibrosis mice. Respir Res 2006;7:134. [56] Meyer M, Huaux F, Gavilanes X, et al. Azithromycin reduces exag- gerated cytokine production by M1 alveolar macrophages in cystic fibrosis. Am J Respir Cell Mol Biol 2009;41:590–602. [57] Bruscia EM, Zhang PX, Ferreira E, et al. Macrophages directly contribute to the exaggerated inflammatory response in cystic fibrosis transmembrane conductance regulator−/− mice. Am J Respir Cell Mol Biol 2009;40:295–304. [58] Gu Y, Harley IT, Henderson LB, et al. Identification of IFRD1 as a
  17. 17. S168 M. Wilke et al. / Journal of Cystic Fibrosis Volume 10 Suppl 2 (2011) S152–S171 modifier gene for cystic fibrosis lung disease. Nature 2009;458:1039– 42. [59] Kulka M, Gilchrist M, Duszyk M, Befus AD. Expression and functional characterization of CFTR in mast cells. J Leukoc Biol 2002;71:54–64. [60] Norkina O, Kaur S, Ziemer D, De Lisle RC. Inflammation of the cystic fibrosis mouse small intestine. Am J Physiol Gastrointest Liver Physiol 2004;286:G1032–41. [61] Xu Y, Tertilt C, Krause A, Quadri LE, Crystal RG, Worgall S. Influence of the cystic fibrosis transmembrane conductance regulator on expression of lipid metabolism-related genes in dendritic cells. Respir Res 2009;10:26. [62] Bonvin E, Le Rouzic P, Bernaudin JF, et al. Congenital tracheal malformation in cystic fibrosis transmembrane conductance regulator- deficient mice. J Physiol 2008;586:3231–43. [63] Liang H, Yang L, Ma T, Zhao Y. Functional expression of cystic fibrosis transmembrane conductance regulator in mouse chondrocytes. Clin Exp Pharmacol Physiol 2010;37:506–8. [64] Rock JR, O’Neal WK, Gabriel SE, et al. Transmembrane protein 16A (TMEM16A) is a Ca2+-regulated Cl secretory channel in mouse airways. J Biol Chem 2009;284:14875–80. [65] Ousingsawat J, Martins JR, Schreiber R, Rock JR, Harfe BD, Kun- zelmann K. Loss of TMEM16A causes a defect in epithelial Ca2+ - dependent chloride transport. J Biol Chem 2009;284:28698–703. [66] Huang F, Rock JR, Harfe BD, et al. Studies on expression and function of the TMEM16A calcium-activated chloride channel. Proc Natl Acad Sci U S A 2009;106:21413–8. [67] Dif F, Marty C, Baudoin C, de Vernejoul MC, Levi G. Severe osteopenia in CFTR-null mice. Bone 2004;35:595–603. [68] Pashuck TD, Franz SE, Altman MK, et al. Murine model for cystic fibrosis bone disease demonstrates osteopenia and sex-related differ- ences in bone formation. Pediatr Res 2009;65:311–6. [69] Haston CK, Li W, Li A, Lafleur M, Henderson JE. Persistent osteope- nia in adult cystic fibrosis transmembrane conductance regulator- deficient mice. Am J Respir Crit Care Med 2008;177:309–15. [70] Bronckers A, Kalogeraki L, Jorna HJ, et al. The cystic fibrosis trans- membrane conductance regulator (CFTR) is expressed in maturation stage ameloblasts, odontoblasts and bone cells. Bone 2010;46:1188– 96. [71] Weinert S, Jabs S, Supanchart C, et al. Lysosomal pathology and osteopetrosis upon loss of H+ driven lysosomal Cl accumulation. Science 2010;328:1401–3. [72] Kornak U, Kasper D, Bosl MR, et al. Loss of the ClC-7 chloride channel leads to osteopetrosis in mice and man. Cell 2001;104:205– 15. [73] Paccou J, Zeboulon N, Combescure C, Gossec L, Cortet B. The prevalence of osteoporosis, osteopenia, and fractures among adults with cystic fibrosis: a systematic literature review with meta-analysis. Calcif Tissue Int 2010;86:1–7. [74] Sermet-Gaudelus I, Castanet M, Retsch-Bogart G, Aris RM. Update on cystic fibrosis-related bone disease: a special focus on children. Paediatr Respir Rev 2009;10:134–42. [75] Hilliard TN, Zhu J, Farley R, et al. Nasal abnormalities in cystic fibrosis mice independent of infection and inflammation. Am J Respir Cell Mol Biol 2008;39:19–25. [76] Grubb BR, Rogers TD, Kulaga HM, et al. Olfactory epithelia exhibit progressive functional and morphological defects in CF mice. Am J Physiol Cell Physiol 2007;293:C574–83. [77] De Jonge HR, Ballmann M, Veeze H, et al. Ex vivo CF diagnosis by intestinal current measurements (ICM) in small aperture, circulating Ussing chambers. J Cyst Fibros 2004;3(Suppl 2):159–63. [78] Boucher RC. Cystic fibrosis: a disease of vulnerability to airway surface dehydration. Trends Mol Med 2007;13:231–40. [79] Elsaesser R, Paysan J. Morituri te salutant? Olfactory signal transduc- tion and the role of phosphoinositides. J Neurocytol 2005;34:97–116. [80] Elsaesser R, Montani G, Tirindelli R, Paysan J. Phosphatidyl-inositide signalling proteins in a novel class of sensory cells in the mammalian olfactory epithelium. Eur J Neurosci 2005;21:2692–700. [81] Ianowski JP, Choi JY, Wine JJ, Hanrahan JW. Substance P stimu- lates CFTR-dependent fluid secretion by mouse tracheal submucosal glands. Pflugers Arch 2008;457:529–37. [82] Ianowski JP, Choi JY, Wine JJ, Hanrahan JW. Mucus secretion by single tracheal submucosal glands from normal and cystic fibro- sis transmembrane conductance regulator knockout mice. J Physiol 2007;580:301–14. [83] Grubb BR, Jones JH, Boucher RC. Mucociliary transport determined by in vivo microdialysis in the airways of normal and CF mice. Am J Physiol Lung Cell Mol Physiol 2004;286:L588–95. [84] Zahm JM, Gaillard D, Dupuit F, et al. Early alterations in airway mucociliary clearance and inflammation of the lamina propria in CF mice. Am J Physiol Cell Physiol 1997;272:C853–9. [85] MacVinish LJ, Gill DR, Hyde SC, et al. Chloride secretion in the trachea of null cystic fibrosis mice: the effects of transfection with pTrial10-CFTR2. J Physiol 1997;499:677–87. [86] Schroeder BC, Cheng T, Jan YN, Jan LY. Expression cloning of TMEM16A as a calcium-activated chloride channel subunit. Cell 2008;134:1019–29. [87] Caputo A, Caci E, Ferrera L, et al. TMEM16A, a membrane protein associated with calcium-dependent chloride channel activity. Science 2008;322:590–4. [88] Yang YD, Cho H, Koo JY, et al. TMEM16A confers receptor-activated calcium-dependent chloride conductance. Nature 2008;455:1210–5. [89] Rochelle LG, Li DC, Ye H, Lee E, Talbot CR, Boucher RC. Distribu- tion of ion transport mRNAs throughout murine nose and lung. Am J Physiol Lung Cell Mol Physiol 2000;279:L14–24. [90] Evans CM, Williams OW, Tuvim MJ, et al. Mucin is produced by clara cells in the proximal airways of antigen-challenged mice. Am J Respir Cell Mol Biol 2004;31:382–94. [91] Park KS, Korfhagen TR, Bruno MD, et al. SPDEF regulates goblet cell hyperplasia in the airway epithelium. J Clin Invest 2007;117:978– 88. [92] Davis CW, Dickey BF. Regulated airway goblet cell mucin secretion. Annu Rev Physiol 2008;70:487–512. [93] Kulaksiz H, Schmid A, Honscheid M, Ramaswamy A, Cetin Y. Clara cell impact in air-side activation of CFTR in small pulmonary airways. Proc Natl Acad Sci U S A 2002;99:6796–801. [94] Chinet TC, Gabriel SE, Penland CM, et al. CFTR-like chloride channels in non-ciliated bronchiolar epithelial (Clara) cells. Biochem Biophys Res Commun 1997;230:470–5. [95] Tiddens HA, Donaldson SH, Rosenfeld M, Pare PD. Cystic fibro- sis lung disease starts in the small airways: can we treat it more effectively? Pediatr Pulmonol 2010;45:107–17. [96] Stick SM, Brennan S, Murray C, et al. Bronchiectasis in infants and preschool children diagnosed with cystic fibrosis after newborn screening. J Pediatr 2009;155:623–8 e1. [97] Fang X, Song Y, Hirsch J, et al. Contribution of CFTR to apical- basolateral fluid transport in cultured human alveolar epithelial type II cells. Am J Physiol Lung Cell Mol Physiol 2006;290:L242–9. [98] Matthay MA, Robriquet L, Fang X. Alveolar epithelium: role in lung fluid balance and acute lung injury. Proc Am Thorac Soc 2005;2:206– 13. [99] Lindert J, Perlman CE, Parthasarathi K, Bhattacharya J. Chloride- dependent secretion of alveolar wall liquid determined by optical- sectioning microscopy. Am J Respir Cell Mol Biol 2007;36:688–96. [100] Thomas GR, Costelloe EA, Lunn DP, et al. G551D cystic fibro- sis mice exhibit abnormal regulation of inflammation in lungs and macrophages. J Immunol 2000;164:3870–7. [101] Kent G, Iles R, Bear CE, et al. Lung disease in mice with cystic fibrosis. J Clin Invest 1997;100:3060–9. [102] Teichgraber V, Ulrich M, Endlich N, et al. Ceramide accumulation mediates inflammation, cell death and infection susceptibility in cystic fibrosis. Nat Med 2008;14:382–91. [103] Guilbault C, De Sanctis JB, Wojewodka G, et al. Fenretinide corrects newly found ceramide deficiency in cystic fibrosis. Am J Respir Cell Mol Biol 2008;38:47–56. [104] Gavilanes X, Huaux F, Meyer M, et al. Azithromycin fails to re-
  18. 18. M. Wilke et al. / Journal of Cystic Fibrosis Volume 10 Suppl 2 (2011) S152–S171 S169 duce increased expression of neutrophil-related cytokines in primary- cultured epithelial cells from cystic fibrosis mice. J Cyst Fibros 2009;8:203–10. [105] van Heeckeren AM, Schluchter MD, Xue W, Davis PB. Response to acute lung infection with mucoid Pseudomonas aeruginosa in cystic fibrosis mice. Am J Respir Crit Care Med 2006;173:288–96. [106] Heeckeren A, Walenga R, Konstan MW, Bonfield T, Davis PB, Ferkol T. Excessive inflammatory response of cystic fibrosis mice to bronchopulmonary infection with Pseudomonas aeruginosa. J Clin Invest 1997;100:2810–5. [107] Davidson DJ, Dorin JR, McLachlan G, et al. Lung disease in the cystic fibrosis mouse exposed to bacterial pathogens. Nat Genet 1995;9:351–7. [108] Stotland PK, Radzioch D, Stevenson MM. Mouse models of chronic lung infection with Pseudomonas aeruginosa: models for the study of cystic fibrosis. Pediatr Pulmonol 2000;30:413–24. [109] McMorran BJ, Palmer JS, Lunn DP, et al. G551D CF mice display an abnormal host response and have impaired clearance of Pseudomonas lung disease. Am J Physiol Lung Cell Mol Physiol 2001;281:L740–7. [110] Oceandy D, McMorran BJ, Smith SN, et al. Gene complementa- tion of airway epithelium in the cystic fibrosis mouse is necessary and sufficient to correct the pathogen clearance and inflammatory abnormalities. Hum Mol Genet 2002;11:1059–67. [111] Bragonzi A, Paroni M, Nonis A, et al. Pseudomonas aeruginosa microevolution during cystic fibrosis lung infection establishes clones with adapted virulence. Am J Respir Crit Care Med 2009;180:138–45. [112] Moser C, Van Gennip M, Bjarnsholt T, et al. Novel experimental Pseudomonas aeruginosa lung infection model mimicking long-term host-pathogen interactions in cystic fibrosis. APMIS 2009;117:95– 107. [113] van Heeckeren AM, Schluchter MD, Drumm ML, Davis PB. Role of Cftr genotype in the response to chronic Pseudomonas aerugi- nosa lung infection in mice. Am J Physiol Lung Cell Mol Physiol 2004;287:L944–52. [114] van Heeckeren AM, Schluchter M, Xue L, et al. Nutritional effects on host response to lung infections with mucoid Pseudomonas aeruginosa in mice. Infect Immun 2004;72:1479–86. [115] Tsai WC, Hershenson MB, Zhou Y, Sajjan U. Azithromycin increases survival and reduces lung inflammation in cystic fibrosis mice. In- flamm Res 2009;58:491–501. [116] Schroeder TH, Reiniger N, Meluleni G, Grout M, Coleman FT, Pier GB. Transgenic cystic fibrosis mice exhibit reduced early clearance of Pseudomonas aeruginosa from the respiratory tract. J Immunol 2001;166:7410–8. [117] Coleman FT, Mueschenborn S, Meluleni G, et al. Hypersusceptibility of cystic fibrosis mice to chronic Pseudomonas aeruginosa oropha- ryngeal colonization and lung infection. Proc Natl Acad Sci U S A 2003;100:1949–54. [118] Reiniger N, Lee MM, Coleman FT, Ray C, Golan DE, Pier GB. Resistance to Pseudomonas aeruginosa chronic lung infection re- quires cystic fibrosis transmembrane conductance regulator-modulated interleukin-1 (IL-1) release and signaling through the IL-1 receptor. Infect Immun 2007;75:1598–608. [119] Raoust E, Balloy V, Garcia-Verdugo I, Touqui L, Ramphal R, Chig- nard M. Pseudomonas aeruginosa LPS or flagellin are sufficient to activate TLR-dependent signaling in murine alveolar macrophages and airway epithelial cells. PloS One 2009;4:e7259. [120] Morris AE, Liggitt HD, Hawn TR, Skerrett SJ. Role of Toll-like receptor 5 in the innate immune response to acute P. aeruginosa pneumonia. Am J Physiol Lung Cell Mol Physiol 2009;297:L1112–9. [121] Quinton PM. Cystic fibrosis: impaired bicarbonate secretion and mu- coviscidosis. Lancet 2008;372:415–7. [122] Cressman VL, Hicks EM, Funkhouser WK, Backlund DC, Koller BH. The relationship of chronic mucin secretion to airway disease in normal and CFTR-deficient mice. Am J Respir Cell Mol Biol 1998;19:853–66. [123] Dif F, Wu YZ, Burgel PR, et al. Critical role of cytosolic phospholi- pase a2α in bronchial mucus hyper-secretion in CFTR-deficient mice. Eur Respir J 2010 Apr 22. [Epub ahead of print] [124] Freedman SD, Katz MH, Parker EM, Laposata M, Urman MY, Al- varez JG. A membrane lipid imbalance plays a role in the phenotypic expression of cystic fibrosis in cftr−/− mice. Proc Natl Acad Sci U S A 1999;96:13995–4000. [125] Freedman SD, Shea JC, Blanco PG, Alvarez JG. Fatty acids in cystic fibrosis. Curr Opin Pulm Med 2000;6:530–2. [126] Beharry S, Ackerley C, Corey M, et al. Long-term docosahexaenoic acid therapy in a congenic murine model of cystic fibrosis. Am J Physiol Gastrointest Liver Physiol 2007;292:G839–48. [127] Guilbault C, Wojewodka G, Saeed Z, et al. Cystic fibrosis fatty acid imbalance is linked to ceramide deficiency and corrected by fenretinide. Am J Respir Cell Mol Biol 2009;41:100–6. [128] Saeed Z, Guilbault C, De Sanctis JB, et al. Fenretinide prevents the development of osteoporosis in Cftr-KO mice. J Cyst Fibros 2008;7:222–30. [129] Harmon GS, Dumlao DS, Ng DT, et al. Pharmacological correction of a defect in PPAR-γ signaling ameliorates disease severity in Cftr- deficient mice. Nat Med 2010;16:313–8. [130] Carvalho-Oliveira IM, Charro N, Aarbiou J, et al. Proteomic analysis of naphthalene-induced airway epithelial injury and repair in a cystic fibrosis mouse model. J Proteome Res 2009;8:3606–16. [131] Borot F, Vieu DL, Faure G, et al. Eicosanoid release is increased by membrane destabilization and CFTR inhibition in Calu-3 cells. PloS One 2009;4:e7116. [132] Becker KA, Riethmuller J, Luth A, Döring G, Kleuser B, Gulbins E. Acid sphingomyelinase inhibitors normalize pulmonary ceramide and inflammation in cystic fibrosis. Am J Respir Cell Mol Biol 2010;42:716–24 [133] Riethmüller J, Anthonysamy J, Serra E, Schwab M, Döring G, Gul- bins E. Therapeutic efficacy and safety of amitriptyline in patients with cystic fibrosis. Cell Physiol Biochem 2009;24:65–72. [134] Becq F, Mall MA, Sheppard DN, Conese M, Zegarra-Moran O. Pharmacological therapy for cystic fibrosis: From bench to bedside. J Cyst Fibros 2011;10(Suppl 2):S129–45. [135] Conese M, Ascenzioni F, Boyd AC, et al. Gene and cell therapy for cystic fibrosis: From bench to bedside. J Cyst Fibros 2011;10(Suppl 2):129–45. [136] Sheppard DN, Welsh MJ. Structure and function of the CFTR chloride channel. Physiol Rev 1999;79(1 Suppl):S23–45. [137] Lansdell KA, Delaney SJ, Lunn DP, Thomson SA, Sheppard DN, Wainwright BJ. Comparison of the gating behaviour of human and murine cystic fibrosis transmembrane conductance regulator Cl chan- nels expressed in mammalian cells. J Physiol 1998;508:379–92. [138] Lansdell KA, Kidd JF, Delaney SJ, Wainwright BJ, Sheppard DN. Regulation of murine cystic fibrosis transmembrane conductance reg- ulator Cl channels expressed in Chinese hamster ovary cells. J Physiol 1998;512:751–64. [139] Van Goor F, Hadida S, Grootenhuis PD, et al. Rescue of CF airway epithelial cell function in vitro by a CFTR potentiator, VX-770. Proc Natl Acad Sci U S A 2009;106:18825–30. [140] De Jonge H, Wilke M, Bot A, Jorna H, Wang W, Kirk K, et al. Differential response of mouse versus human CFTR to CFTR potentiators. Pediatr Pulmonol Suppl 2007;30:208. [141] Scott-Ward TS, Cai Z, Dawson ES, et al. Chimeric constructs endow the human CFTR Cl channel with the gating behavior of murine CFTR. Proc Natl Acad Sci U S A 2007;104:16365–70. [142] Noël S, Wilke M, Bot AG, De Jonge HR, Becq F. Parallel im- provement of sodium and chloride transport defects by miglustat (n-butyldeoxynojyrimicin) in cystic fibrosis epithelial cells. J Pharma- col Exp Ther 2008;325:1016–23. [143] Norez C, Noel S, Wilke M, et al. Rescue of functional delF508- CFTR channels in cystic fibrosis epithelial cells by the α-glucosidase inhibitor miglustat. FEBS Lett 2006;580:2081–6. [144] Lubamba B, Lecourt H, Lebacq J, et al. Preclinical evidence that sildenafil and vardenafil activate chloride transport in cystic fibrosis. Am J Respir Crit Care Med 2008;177:506–15.