- Email: [email protected]
Mouse models of cystic fibrosis
TRENDS in Genetics, Vol.17 No.10, October 2001
A TRENDS Guide to Mouse Models of Human Diseases
|
Review
Mouse models of cystic fibrosis Donald J. Davidson and Mark Rolfe The development of mouse models for cystic fibrosis has provided the opportunity to dissect disease pathogenesis, correlate genotype and phenotype, study disease-modifying genes and develop novel therapeutics. This review discusses the successes and the challenges encountered in characterizing and optimizing these models. Cystic fibrosis (CF) is a common, lethal, autosomal recessive disorder caused by mutations in the CFTR gene, with the most common mutation (∆F508) occurring on ~70% of CF chromosomes. Dysfunction of the CFTR protein, which acts as an apically localized epithelial chloride ion channel, results in the classical manifestations of CF: salty sweat, pancreatic insufficiency, intestinal obstruction, male infertility, and severe pulmonary disease, with characteristic abnormalities in electrolyte transport. The most serious consequence is progressive and ultimately fatal inflammatory lung disease characterized by chronic microbial colonization and repeated acute exacerbations of pulmonary infection, with a distinctive spectrum of pathogens.These clinical manifestations show considerable variation between individuals because of an as yet incompletely understood combination of environmental factors, independently segregating disease-modifying genes, and differences between specific CFTR mutations. A combination of chest physiotherapy, aggressive antibiotic therapy and pancreatic enzyme supplementation constitutes the traditional mainstay of treatment for CF. Despite the success of these therapies, novel approaches are required, both to achieve further increases in life expectancy and to improve quality of life rather than simply to alleviate symptoms. As our understanding of the molecular and biological basis of CF becomes more comprehensive, so the goal of successful treatments becomes more accessible. Generating mouse models of cystic fibrosis The first mouse models of CF were created within three years of the isolation of the human CFTR gene. To date, 11 different mouse models of CF have been characterized (Table 1). These models can be categorized as those designed simply to disrupt Cftr expression and those that specifically model various human clinical mutations. The former can be further subdivided into those that used a replacement strategy to disrupt the Cftr gene, creating absolute nulls (with no normal CFTR production), and those that used insertion into the target gene without loss of genomic sequence, in which the mutants retain the
potential to produce low levels of normal Cftr mRNA by various mechanisms.These distinctions affect the phenotype of the models and must be recognized in any interpretation of the phenotypes observed.
In association with MKMD
Phenotypes of mouse models of cystic fibrosis Characterization of the different mouse models of CF has demonstrated most of the same primary phenotypes, including intestinal obstruction, reduced fertility and characteristic intestinal and airway electrophysiology, reproducing many of the manifestations of CF in humans. Important differences have been observed, however, both between different models (particularly in survival rates), and between mouse models of CF and human disease patterns (most clearly observed in pancreatic function and the murine pulmonary phenotype). The phenotypic variations between mouse models have been shown to relate to the specific mutation of Cftr generated, to environmental influences and to independently segregating modifier genes. Given rigorous evaluation, these differences provide the means to start dissecting the key components in disease pathogenesis; however, the need to clearly define the mouse model studied becomes obvious. In addition, the most robust phenotypes have provided in vivo models for the development and optimization of novel therapeutics.
http://research.bmn.com/mkmd
Survival and intestinal disease Intestinal pathology and the resultant mortality are the hallmarks of Cftr mutation in the mouse. Considerable variation in the specific pathology and the degree of severity has been reported between different models. In most cases, however, characterization of the mutant mice has revealed abnormal electrophysiological profiles, runting and failure to thrive, goblet cell hyperplasia, mucin accumulation, crypt dilation and intestinal obstruction (bearing similarity to meconium ileus, present in 10–15% of CF patients) with resultant perforation, peritonitis and death. Studies characterizing the electrophysiological profile have found broadly similar phenotypes in the different models (Table 2) with a significant decrease in the baseline potential difference (probably representing a decreased
0168-9525/01/$ – see front matter © 2001 Elsevier Science Ltd. All rights reserved. PII: S0168-9525(01)02452-0
Donald J. Davidson* University of British Columbia, BC Research Institute for Child and Family Health, Room 381, 950 West 28th Avenue, Vancouver, British Columbia, Canada V5Z 4H4. *e-mail: [email protected]
Mark Rolfe Medical Research Council Human Genetics Unit, Western General Hospital, Crewe Road, Edinburgh, UK EH4 2XU.
S29
Review
|
A TRENDS Guide to Mouse Models of Human Diseases
TRENDS in Genetics, Vol.17 No.10, October 2001
Table 1. Mouse models of cystic fibrosisa Mouse
Mutation
Cftr mRNA
Original strain
Cftrtm1Unc
(D) Exon 10 replacement
No wildtype (wt) mRNA detectable
C57Bl/6/129, BALB/c/129, B6D2/129
Refs 39
Cftrtm1Hgu
(D) Exon 10 insertional
10% of normal levels of wt mRNA
MF1/129
40
Cftrtm1Cam
(D) Exon 10 replacement
No wt mRNA detectable
MF1/129, C57Bl/6/129
41
Cftrtm1Hsc
(D) Exon 1 replacement
No wt mRNA detectable
CD1/129
3
Cftrtm1Bay
(D) Exon 3 insertional duplication
<2% of normal levels of wt mRNA
C57Bl/6/129
42
Cftrtm3Bay
(D) Exon 2 replacement
No wt mRNA detectable
C57Bl/6/129
43
Cftrtm2Cam
(C) ∆F508 Exon 10 replacement
Mutant mRNA 30% of normal expression levels
C57Bl/6/129
44
Cftrtm1Kth
(C) ∆F508 Exon 10 replacement
Decrease in mutant mRNA levels in intestinal tract
C57Bl/6/129
45
Cftrtm1Eur
(C) ∆F508 Exon 10 insertional 'hit and run’
Mutant mRNA expression at normal levels
FVB /129
46
Cftrtm1G551D
(C) G551D Exon 11 replacement
Mutant mRNA 53% of normal expression levels
CD1/129
6
Cftrtm2Hgu
(C) G480C Exon 10 insertional 'hit and run'
Mutant mRNA expression at normal levels
C57Bl/6/129
b
a
Mouse models of cystic fibrosis (CF) can be categorized as those that simply disrupt Cftr expression (D), and those that model specific clinical mutations (C). b Dickinson, P. et al., Generation of a CF mutant mouse possessing the G480C mutation. 22nd European Cystic Fibrosis Conference, 13–19 June 1998, Berlin.
rate of unstimulated Cl− secretion) and a complete absence, or a significant decrease, in cAMP-stimulated Cl− secretion (indicative of the loss of CFTR function). These profiles can be used to distinguish unequivocally mutants from their wild-type (wt) littermates and closely model the electrophysiological phenotype in CF humans. The survival rates of the different mouse models generally reflect the severity of intestinal pathology and vary from <5% in Cftrtm1Unc/Cftrtm1Unc nulls, and similar rates in the majority of models, to ~90% in Cftrtm1Hgu/Cftrtm1Hgu mice, and normal survival in Cftrtm2Hgu/Cftrtm2Hgu and Cftrtm1Eur/Cftrtm1Eur mice (Table 3). These dramatic differences appear to be the result of mutation-specific effects, independently segregating modifier genes and environmental influences. A low-level production of ~10% of normal CFTR has been proposed to be the explanation for the significantly greater survival rate in the residual function Cftrtm1Hgu/Cftrtm1Hgu mice1. In addition, in studies using mutant allele crosses, compound heterozygote mice have indicated that small levels of normal CFTR activity can have dramatic effects on survival2. These important studies suggest that gene therapy and other replacement or augmentation therapeutic strategies that are able to provide even modest restoration of function could provide significant benefits in CF individuals.The reason for the mild nature of the intestinal disease in the Cftrtm2Hgu/Cftrtm2Hgu and Cftrtm1Eur/Cftrtm1Eur mice is less clear. However, it could relate to the use of a double homologous recombination (‘hit and run’) strategy to generate these mice. As a result, both these mice
S30
express normal levels of a mutant CFTR protein, which could provide enough residual function in mice to change the phenotype from that observed in a ‘null’ animal or one expressing a low level of mutant protein. In studies using Cftrtm1Hsc/Cftrtm1Hsc mice bred to congenicity on different inbred backgrounds, the mortality has been shown to manifest at two distinct periods and to be partially determined by an independently segregating modifier locus3. These studies led to the identification of a genetic modifier locus for meconium ileus in humans4. This powerful technique to find disease-modifying genes is made possible in mice by the use of lines of genetically identical inbred animals. An important benefit of murine models of disease, this has the potential to provide great insights into the variability of disease manifestation and reveal possible targets for novel therapeutic approaches. Finally, survival has been shown to be influenced both by diet and housing conditions.The use of a liquid diet has been found to prolong the lifespan of Cftrtm1Unc/Cftrtm1Unc mice5, whereas the mortality observed in Cftrtm1G551D/ Cftrtm1G551D mice has been shown to be partially dependent upon the sterility of their housing conditions6. Analyses of these variable phenotypes has provided insights into the role of CFTR in the intestinal tract in vivo and demonstrated that this severe phenotype can be alleviated in mice by a low level of normal CFTR, normal levels of mutant CFTR or the presence of certain modifier genes. Although rather more severe than in humans, the intestinal phenotype in these mouse models is sufficiently
TRENDS in Genetics, Vol.17 No.10, October 2001
A TRENDS Guide to Mouse Models of Human Diseases
|
Review
Table 2. Intestinal electrophysiology in mouse models of cystic fibrosisa Mutation
Tissue
Baseline PD
cAMP-mediated Cl– response
Ca2+-related Cl– response
Refs
CF human
GI tract
↔ or ↑
↓
↓
Cftrtm1Unc
Jejunum
↓
↓ 100%
↓
47
Caecum
↓
↓ 100%
↓
48
Colon
↓
↓ 100%
↓
48
Jejunum
↓
↓ 65%
↓
49
Cftrtm1Hgu
Caecum
↓
↓ 65%
↓
49
Cftrtm1Cam
Caecum
↓
↓ 100%
n.r.
41
Cftrtm1Hsc
Rectum
n.r.
↓ 100%
↑
3
Ileumb
n.r.
↓ 100%
↑
3
Cftrtm1Bay
Ileum
↔
↓ 80%
n.r.
42
Cftrtm3Bay
Colon
n.r.
↓ 100%
↓
43
Cftrtm2Cam
Colon
↓
↓ 100%
↓
44
Cftrtm1Kth
Jejunum
↔
↓ 100%
n.r.
45
CftrtmEur
Ileum
↓
↓ 66%
↔
46
Caecum
n.r.
↓ 92%
n.r
50
Jejunum
↓
↓ 99%
n.r.
6
Caecum
↓
↓ 95%
n.r.
6
Caecum
↓
↔
↓
c
Cftrtm1G551D Cftrtm2Hgu aComparison
of the electrophysiological profiles of the intestinal epithelium in human cystic fibrosis (CF) and mouse models of CF, on the original background strain. Increased (↑), decreased (↓) or preserved (↔) potential difference (PD) in comparison with non-CF controls. Abbreviations: GI, gastrointestinal tract; n.r., not reported. bPatch-clamped, isolated ileal crypt cells. cDickinson, P. et al., Generation of a CF mutant mouse possessing the G480C mutation. 22nd European Cystic Fibrosis Conference, 13–19 June 1998, Berlin.
similar to suggest the same pathophysiological processes, validating their use as models for human disease and the relevance of these findings for novel therapy development. Pancreatic disease Pancreatic insufficiency is a prominent manifestation of CFTR dysfunction in humans, but has not been convincingly demonstrated in most mouse models of CF.This appears to be the result of low levels of expression of Cftr in the murine pancreas and the presence of an alternative fluid secretory pathway, which is activated by increases in intracellular calcium7.This indicates that other ion channels might be capable of compensating for the loss of CFTR and suggests novel therapeutic approaches in humans, namely to identify and utilize such pathways. One study of Cftrtm1Unc/Cftrtm1Unc mice weaned on a liquid diet to increase survival rates, demonstrated significant differences in pancreatic growth and specific enzyme activities8. Similar, although less severe, abnormalities in wt controls, however, suggested that the abnormalities were predominantly secondary to malnutrition. A further study using a liquid elemental diet reported luminal dilatation
and the accumulation of zymogen granules at the apical pole of the ductal epithelial cells in Cftrtm1Unc/Cftrtm1Unc mice. This phenotype has since been used, and corrected with oral administration of docosahexanoic acid (DHA), in a study of the role of dietary fatty acids in CF (Ref. 9). This study suggested that a primary defect in fatty acid metabolism might play a significant role in the pathogenesis of CF and indicated DHA as a novel therapy. However, the role of the liquid diet in this phenotype might yet prove to be significant. Lung disease Lung disease represents the primary concern in CF and the manifestation for which an animal model is likely to be the most valuable. Although a variety of pulmonary abnormalities have been reported, these are complex, mostly precipitated in response to exposure to pathogens, and do not fully replicate the pathogenesis of lung disease in CF individuals. Nevertheless, the development of lung phenotypes secondary to the loss of CFTR function in mice provides an important in vivo system to study the pathogenesis of CF lung disease.
S31
|
Review
A TRENDS Guide to Mouse Models of Human Diseases
TRENDS in Genetics, Vol.17 No.10, October 2001
Table 3. Survival in mouse models of cystic fibrosisa Mutation
Perinatal death
Survival to maturity
Body weight
∆F508 CF human
10% MI
20% DIOS
Failure to thrive
Cftrtm1Unc null
50% by day 7 40% death at weaning
<5% survival to maturity
10–50% reduction
39
Cftrtm1Hgu residual function
5% by day 7 2% death at weaning
90% survival to maturity
No reduction
40
Cftrtm1Cam null
80% by day 7 10% death at weaning
<5% survival to maturity
50% reduction
41
Cftrtm1Hsc null
55% by day 7 20% death at weaning
25% survival to maturity
Delayed
3
Cftrtm1Bay null
40% by day 7 10% death at weaning
n.r.
70% reduction
42
Cftrtm3Bay null
n.r.
40% survival at one month
Reduced
43
Cftrtm2Cam ∆F508
35% by day 16
<5% survival to maturity
n.r.
44
Cftrtm1Kth ∆F508
10% by day 7
40% survival to maturity
50% reduction
45
Cftrtm1Eur
∆F508
Refs
None
Normal
20% reduction
46
Cftrtm1G551D G551D
n.r.
67% survival at day 35 in SPF conditions 27% survival at day 35 in normal conditions
30–50% reduction
6
Cftrtm2Hgu G480C
None
Normal
No reduction
b
a
Comparisons between the survival rates of mouse models of cystic fibrosis (CF). Abbreviations: DIOS, distal intestinal obstruction syndrome; MI, meconium ileus; n.r., not reported; SPF, specific pathogen free. b Dickinson, P. et al., Generation of a CF mutant mouse possessing the G480C mutation. 22nd European Cystic Fibrosis Conference, 13–19 June 1998, Berlin.
Electrophysiological studies Electrophysiological analyses of nasal and tracheal airway epithelia (Tables 4 and 5) in mouse models of CF have been shown to differentiate clearly between mutants and wt littermates, even in the absence of gross pathology. These studies clearly demonstrate the basic ion-channel defect, with the nasal epithelium of mouse models of CF accurately replicating the human profile. Analysis of the trachea has, however, proved to be more complex, with important differences in both sodium- and chloride-ion transport observed between murine and human profiles. Indeed, it has been proposed that alternative chloride-ion pathways dominate over CFTR in the mouse trachea and might alleviate the effects of CFTR dysfunction10.The extent to which these observations are replicated in the lower airways remains unknown. Analyses of inbred mouse strains have revealed considerable variation and demonstrated that the ion transport properties in the murine airways are regulated by independently segregating modifier genes.Thus, the consequences of CFTR dysfunction in the trachea might vary considerably between different mouse models of CF.This raises the possibility of dissecting out the component parts of the electrophysiological response and establishing their relative contributions to disease pathogenesis. The ability to distinguish, unequivocally, mutant mice from wt littermates using electrophysiological profiles has
S32
been crucial in the use of mouse models of CF for testing the efficacy of novel therapies, particularly somatic gene therapy. The basis of the gene therapy strategy for CF is the prevention of disease development by direct replacement of CFTR gene function, irrespective of a complete understanding of disease pathogenesis. Gene correction strategies must therefore be demonstrated to be safe and effective so that intervention can be attempted in infants. Successful correction towards the wt pulmonary electrophysiological phenotype has provided the primary endpoint for analyses in mouse models of CF, using purified CFTR protein11 and both adenoviral-based12 and liposome-based13,14 gene therapy vectors. These studies were instrumental in the initiation of human gene therapy trials, provided results similar to those achieved in humans, and revealed many of the same obstacles to successful correction. These issues, such as low transfection efficiency, transient expression and effect of repeat administration, are now being addressed. Mouse models continue to prove an important resource in this development and optimization process and bridge the gap between cell-culture-based studies and human trials. Lung pathology The use of electrophysiological profiles has been of great importance in the development of novel therapies; however, disease state endpoints are also vital to evaluate the
TRENDS in Genetics, Vol.17 No.10, October 2001
A TRENDS Guide to Mouse Models of Human Diseases
|
Review
Table 4. Nasal electrophysiology in mouse models of cystic fibrosisa Mutation
Baseline PD
Amiloride response
cAMP-mediated Cl– response
Ca2+-related Cl– response
CF human
↑
↑
↓
↔
Cftrtm1Unc
↑
↑
↓ 100%
↑
51
Cftrtm1Hgu
↑
↑
↓ 70%
↔
49
Cftrtm1Cam
↑
n.r.
n.r.
n.r.
13
Cftrtm1Hsc
↑
↑
↓ 100%
↑
52
Cftrtm1Eur
↑
↑
Response to Cl– gradient
46
Cftrtm1Kth
↑
↑
↓ 100%
n.r.
45
Cftrtm1G551D
↑
↑
↓ 100%
↔
6
Refs
aComparison
of the electrophysiological profiles of the nasal epithelium in human cystic fibrosis (CF) and mouse models of CF, on the original mixed genetic background strains. Profiles are shown as increased (↑), decreased (↓) or preserved (↔) potential difference (PD) in comparison with non-CF controls. Amiloride inhibits the epithelial sodium channel EnaC. Abbreviation: n.r., not reported.
pathophysiological consequences. These have proved to be considerably more complex (Table 6). The first observations relating to abnormal lung pathology in mouse models of CF were made in outbred MF1/129 Cftrtm1Hgu/Cftrtm1Hgu mice. No gross lung disease was observed at birth, or in animals born and raised in isolator conditions15, but cytokine abnormalities were observed in mutant mice maintained in standard animal facilities16.These observations and the studies that followed suggest that an abnormal lung phenotype might not manifest without exposure to pathogens. In response to exposure to aerosolized clinical isolates of Staphylococcus aureus and Burkholderia cepacia, Cftrtm1Hgu/Cftrtm1Hgu mice demonstrated significantly impaired airway clearance and the
development of significantly more severe, pathogen-specific, lung pathology15.These observations in response to clinically relevant B. cepacia infection have recently been repeated and elaborated upon using Cftrtm1Unc mice maintained on a liquid diet17. This recent model also demonstrates an increased influx but suboptimal activation of inflammatory cells in the lungs of the mutant mice. In earlier studies, however, abnormal lung pathology was not observed in Cftrtm1Unc mice following exposure to S. aureus (Ref. 18). It is probable that mutation-specific effects, differences in the mouse-strain backgrounds (particularly variations in the role of alternative airway epithelial ion-channels) and varied methods of bacterial exposure influence these contrasting observations.
Table 5. Tracheal electrophysiology in mouse models of cystic fibrosisa Amiloride response
cAMP-mediated Cl– response
Ca2+-related Cl– response
Mutation
Baseline PD
CF human
↑ or ↔
↑
↓
↔
Cftrtm1Unc
↔
↔
↔
↔
10
Cftrtm1Hgu
↓
↓
↓ 60%
↔
49
Cftrtm1Cam
↓
↓
↓ 75%
↔
13
Cftrtm1Bay
↔
n.r.
↓b 70%
n.r.
42
Cftrtm2Cam
↔
↔
↔c
Cftrtm1G551D
↔
↔
↓ 60%
to ↓ 60%
Refs
↑
44
↑
6
a
Comparison of the electrophysiological profiles of the tracheal epithelium in human cystic fibrosis (CF) and mouse models of CF on the original mixed genetic background strains. Increased (↑), decreased (↓) or preserved (↔) potential difference (PD) in comparison to non-CF controls. Amiloride inhibits the epithelial sodium channel ENaC. Abbreviation: n.r., not reported. b Greatest decrease observed in the youngest mice. c Studied in cultured foetal tracheal cells.
S33
Review
|
A TRENDS Guide to Mouse Models of Human Diseases
TRENDS in Genetics, Vol.17 No.10, October 2001
Table 6. Lung phenotypes in mouse models of cystic fibrosisa Mutation
Observation
Refs
Cftrtm1Hgu
↓ Pulmonary clearance of Staphylococcus aureus and Burkholderia cepacia
15
↑ Pulmonary pathology after repeated exposure to S. aureus and B. cepacia
15
↑ Bronchoalveolar lavage tumour necrosis factor α
16
↑ Inflammatory cells in tracheal lamina propria
19
↓ Tracheal mucociliary transport
19
↔ Pulmonary clearance of Pseudomonas aeruginosa
b
Altered submucosal gland distribution
20
Cftrtm1G551D
↑ S100 Ca
53
Cftrtm1Unc
↓ Pulmonary clearance and ↑ pulmonary pathology after repeated exposure to B. cepacia 17
Cftrtm1Kth
2+
binding protein and tumour necrosis factor α
↔ Pulmonary clearance of P. aeruginosa and S. aureus
18
↓ Airway mucociliary transport
21
↑ Mortality from P. aeruginosa-laden agar beads
25
Cytokine abnormalities
24
↓ iNOS expression
29
↔ Pulmonary clearance of P. aeruginosa
22
↓ Lung cell ingestion, ↑ lung burden with P. aeruginosa
23
↓ iNOS expression
28
aComparison
of the pulmonary pathology observed in different mouse models of cystic fibrosis (CF). Observations: ↑, increased; ↓, decreased; ↔, no change in comparison to non-CF controls. bLarbig, M. et al., Pseudomonas aeruginosa infection in cystic fibrosis: animal model of the transgenic cftrm1HGU mouse. European Cystic Fibrosis Conference, 13–19 June 1998, Berlin.
Further studies using Cftrtm1Hgu mice have demonstrated significantly impaired mucociliary transport of inert particles in vivo (Ref. 19) and an altered distribution of mucus and antimicrobial airway surface liquid (ASL)-secreting submucosal glands20 suggesting possible mechanisms for disease development, secondary to CFTR dysfunction. A subsequent study using embedded, cultured lung slices from mice homozygous for the Cftrtm1Unc mutation, partially backcrossed onto the C57Bl/6 background, also demonstrated impaired mucociliary transport21. Although these studies provide evidence of an abnormal response to bacterial burden, initial studies examining the effect of exposure to the classic lung pathogen Pseudomonas aeruginosa failed to demonstrate any abnormal lung phenotype in Cftrtm1Hgu/Cftrtm1Hgu mice (Larbig, M. et al., Pseudomonas aeruginosa infection in cystic fibrosis: animal model of the transgenic cftrm1HGU mouse. European Cystic Fibrosis conference, 13–19 June 1998, Berlin), Cftrtm1Unc/Cftrtm1Unc mice18 or Cftrtm1Kth/Cftrtm1Kth mice (carrying the ∆F508 mutation)22. These studies suggested that, despite abnormal responses to other bacteria, mouse models of CF might not display increased susceptibility to pulmonary infection with P. aeruginosa. This is obviously a significant contrast to CF lung disease in humans, the explanation for which remains elusive. However, a recent study examining the in vivo significance of epithelial cell internalization of P. aeruginosa
S34
quantified and localized these organisms within a few hours after delivery23. Under the conditions of this study, significant differences were observed between various mouse models of CF and controls, suggesting that basic defects in the host interaction with this organism might indeed exist in the mouse models, despite the elusive nature of convincing pathology. The study of P. aeruginosa in mouse models of CF is probably complicated by the phenotypic alteration that this organism undergoes over the course of chronic infection of the lungs of CF individuals. Although the initial infections in CF individuals are with planktonic strains, rapid deterioration of the CF lung usually occurs after transformation of P. aeruginosa into the mucoid form in the host. It is therefore unclear whether the most revealing studies will arise from trying to mimic this process by challenging mice with non-mucoid P. aeruginosa to evaluate predisposition to infection, or modeling colonization with mucoid strains in the absence of previous rounds of infection. In addition, the significance of previous infections with other organisms, and the consequent antibacterial chemotherapy received, is unclear with respect to priming the CF lung for P. aeruginosa infection. To date, studies using agar beads laden with P. aeruginosa to model colonization have been more successful than the alternative approach. Using this technique, significantly
TRENDS in Genetics, Vol.17 No.10, October 2001
decreased survival rates have been demonstrated in Cftrtm1Unc/Cftrtm1Unc mice24,25; however, the basis for this observation is unclear. Bacterial proliferation was demonstrated regardless of genotype, only one of the studies recovered significantly higher numbers of bacteria from the lungs of mutant mice, and no significant differences in lung histopathology were observed between mutant mice and non-CF littermates. Perhaps the most significant observations were the significantly higher levels of proinflammatory cytokines and the decreased levels of the anti-inflammatory cytokine IL-10 in Cftrtm1Unc/Cftrtm1Unc mice, despite an absence of significant differences in the bacterial lung burden or pulmonary histopathology24. The role of the liquid diet fed to the mice in these studies could, however, prove to be significant, with a malnourished mouse model reported to demonstrate some striking similarities26. Nevertheless, the agar bead model reveals interesting phenotypic differences between mouse models of CF and control animals, and could prove effective in the study of the host response to established infection. This model has since been used in the further development of adenoviral-mediated gene transfer27. A further phenotype described in mouse models of CF is that of decreased pulmonary levels of the inducible isoform of nitric oxide synthase (iNOS), which has been implicated in the development of CF lung disease.The expression of iNOS is significantly reduced in mixed background strain Cftrtm1Kth/Cftrtm1Kth mice, homozygous for the ∆F508 mutation28, and Cftrtm1Unc/Cftrtm1Unc mice29. Furthermore, in Cftrtm1Unc/Cftrtm1Unc mice expressing human CFTR cDNA in the intestinal tract but not the nose, iNOS expression is observed in the ileum but not in the nasal epithelium29. The exact mechanism by which CFTR affects the expression of iNOS remains to be determined. In conclusion, despite some significant similarities between CF lung disease in humans and mouse models of CF and promising recent developments17,23, significant differences are also evident under the experimental conditions described, and the suitability of these models as endpoints for therapeutic testing remains controversial. Testing hypotheses in mouse models of cystic fibrosis Despite certain differences between the pathology observed in mouse models of CF and human CF individuals, the former clearly demonstrate a range of abnormal phenotypes as a result of Cftr mutation. Consequently, mouse models of CF have been used to perform in vivo analysis of several hypotheses describing the pathogenesis of CF that were generated primarily on the basis of in vitro studies. Recent hypotheses have emphasized the role of CFTR in determining either the volume or the ionic concentration of the ASL lining in the lung epithelia, affecting mucociliary clearance or the antibacterial activity of salt-sensitive peptides30. Contrasting hypotheses were established on the basis of in vitro studies and limited analysis of human
A TRENDS Guide to Mouse Models of Human Diseases
|
Review
ASL. Several studies have circumvented some of the complicating factors by measuring the ionic composition of ASL in airways of various mouse models of CF, using techniques that could not be performed in human subjects31–34. Although the results of these studies suggest that Cftr mutation does not affect ASL salt concentration, they conflict over the in vivo tonicity. However, these studies used contrasting techniques on mouse models of CF with different mutations in Cftr, on different background strains. All these potential influences, and others (e.g. the role and distribution of submucosal glands) can be controlled for, and altered, in future studies.Thus, the crucial factors influencing data from mouse model studies should be more easily addressed in future studies and provide in vivo answers to the questions raised by these hypotheses. In addition, lines of transgenic mice with ‘knockout’ mutations in genes encoding antibacterial peptides have been established and are currently being characterized.The analysis of these mutant mice and the offspring intercrossed with mouse models of CF will help to determine the role of these antibacterial agents in the lung, particularly with reference to CF lung disease. Mouse models of CF have also been used to examine the hypothesis that CFTR is a receptor for P. aeruginosa complete lipopolysaccharide core, resulting in bacterial internalization into epithelial cells. This mechanism has been proposed to perform a protective role in the normal lung and to be compromised in CF (Ref. 35). A recent study examined the host interaction with P. aeruginosa just 4.5 hours after bacterial delivery23.This study demonstrated significantly less ingestion of P. aeruginosa by lung cells and significantly greater bacterial lung burdens in various mouse models of CF compared to wt controls. Confocal and scanning electron microscopic imaging was used to demonstrate association between airway epithelial cells and bacteria in the wt controls, suggesting a possible role for these cells in host defense. By contrast, a study using a range of transgenic mice expressing varying levels of human and murine CFTR found no direct correlation between the level of CFTR expression and the pulmonary clearance of P. aeruginosa, or the association of bacteria with epithelial cells in vivo (Ref. 36). The time points and the choice of strains evaluated in these studies appear to be crucial23, however, emphasizing the requirement for rigorous evaluation in optimization of these phenotypes and perhaps the subtlety of the underlying defect. Mouse models of CF have also been used to examine possible hypotheses for the high frequency of the mutant CF allele in human Caucasian populations.These studies have suggested two possible heterozygote advantage theories. In Cftrtm1Unc/Cftrtm1Unc mice, expressing no Cftr, intestinal fluid secretion was not observed in response to cholera toxin, and in heterozygote mice secretion was reduced to 50% of normal wt levels37. This suggests that individuals carrying a mutant CF allele might be more resistant to the potentially fatal diarrhea and dehydration induced by
S35
Review
|
A TRENDS Guide to Mouse Models of Human Diseases
Vibrio cholerae. A later study demonstrated that homozygous and heterozygous Cftrtm2Cam mice were protected from intestinal epithelial invasion by Salmonella typhii, lending credence to the idea that mutant CF alleles might confer a resistance to typhoid fever38. Thus, the ability to pursue invasive techniques in vivo and to control for genetic and environmental influences in future studies makes mouse models of CF a powerful system to help dissect the pathogenesis of CF lung disease and define many of the critical components of disease development. Future directions The challenges for future use of mouse models of CF initially lie in refining existing models to replicate human disease as accurately as possible and then in understanding the mechanisms that underlie the development of the mutant phenotypes observed only in mice. The phenotypic variability observed in mouse models of CF as a result of different strain backgrounds, specific mutations in Cftr, and environmental influences, is of great potential for the definition of genetic modifiers and to refine the mouse models of CF. These processes should define many of the key components involved in disease pathogenesis, establish the most suitable models for therapy testing with clear relevant clinical endpoints, and suggest novel therapeutic approaches for the treatment of human disease. The future availability of comprehensive gene expression arrays could prove particularly illuminating in studying the mechanisms underlying the phenotypic differences observed between different models, and between mouse models of CF and non-CF littermates, in response to different environmental stimuli. By adopting such an approach, mouse models of CF can provide valuable contributions to our understanding of this disease process and support efforts towards organ-based treatment for CF patients. Acknowledgements We gratefully acknowledge Julia R. Dorin, David J. Porteous, Eric W.F.W. Alton and Steve Smith. Donald Davidson is a Wellcome Trust fellow. Mark Rolfe is supported by the Medical Research Council, UK. References 1 Dorin, J.R. et al. (1994) Long term survival of the exon 10 insertional cystic fibrosis mutant mouse is a consequence of low level residual wild type Cftr gene expression. Mamm. Genome 5, 465–472 2 Dorin, J.R. et al. (1996) A demonstration using mouse models that successful gene therapy for cystic fibrosis requires only partial gene correction. Gene Ther. 3, 797–801 3 Rozmahel, R. et al. (1996) Modulation of disease severity in cystic fibrosis transmembrane conductance regulator deficient mice by a secondary genetic factor. Nat. Genet. 12, 280–287 4 Zielenski, J. et al. (1999) Detection of a cystic fibrosis modifier locus for meconium ileus on human chromosome 19q13. Nat. Genet. 22, 128–129 5 Kent, G. et al. (1996) Phenotypic abnormalities in long-term surviving cystic fibrosis mice. Pediatr. Res. 40, 233–241
S36
TRENDS in Genetics, Vol.17 No.10, October 2001
6 Delaney, S.J. et al. (1996) Cystic fibrosis mice carrying the missense mutation G551D replicate human genotype-phenotype correlations. EMBO J. 15, 955–963 7 Gray, M.A. et al. (1995) Chloride channels and cystic fibrosis of the pancreas. Biosci. Rep. 15, 531–541 8 Ip, W.F. et al. (1996) Exocrine pancreatic alteration in long-lived surviving cystic fibrosis mice. Pediatr. Res. 40, 242–249 9 Freedman, S.D. et al. (1999) A membrane lipid imbalance plays a role in the phenotypic expression of cystic fibrosis in cftr−/− mice. Proc. Natl.Acad. Sci. U. S.A. 96, 13995–14000 10 Grubb, B.R. et al. (1994) Anomalies in ion transport in CF mouse tracheal epithelium. Am. J. Physiol. 267, C293–C300 11 Ramjeesingh, M. et al. (1998) Assessment of the efficacy of an in vivo CFTR protein replacement therapy in CF mice. Hum. Gene Ther. 9, 521–528 12 Grubb, B.R. et al. (1994) Inefficient gene transfer by adenovirus vector to cystic fibrosis airway epithelia of mice and humans. Nature 371, 802–806 13 Hyde, S.C. et al. (1993) Correction of the ion transport defect in cystic fibrosis transgenic mice by gene therapy. Nature 362, 250–255 14 Alton, E.W.F.W. et al. (1993) Non-invasive liposome-mediated gene delivery can correct the ion transport defect in cystic fibrosis mutant mice. Nat. Genet. 5, 135–142 15 Davidson, D.J. et al. (1995) Lung disease in the cystic fibrosis mouse exposed to bacterial pathogens. Nat. Genet. 9, 351–357 16 Morrison, G. et al. (1997) Analysis of the pulmonary inflammatory phenotype in the CF mutant mouse. Pediatr. Pulmonol. (Suppl. 14), 317 17 Sajjan, U. et al. (2001) Enhanced susceptibility to pulmonary infection with Burkholdera cepacia in Cftr−/− mice. Infect. Immun. 69, 5138–5150 18 Cressman,V.L. et al. (1998) The relationship of chronic mucin secretion to airway disease in normal and CFTR-deficient mice. Am. J. Respir. Cell Mol. Biol. 19, 853–866 19 Zahm, J.M. et al. (1997) Early alterations in airway mucociliary clearance and inflammation of the lamina propria in CF mice. Am. J. Physiol. 41, C853–C859 20 Borthwick, D.W. et al. (1999) Murine submucosal glands are clonally derived and show a Cftr dependent distribution pattern. Am. J. Respir. Cell Mol. Biol. 20, 1181–1189 21 Cowley, E.A. et al. (1997) Mucociliary clearance in cystic fibrosis knockout mice infected with Pseudomonas aeruginosa. Eur. Respir. J. 10, 2312–2318 22 McCray, P.B. et al. (1999) Efficient killing of inhaled bacteria in ∆F508 mice: role of airway surface liquid composition. Am. J. Physiol. 277, L183–L190 23 Schroeder,T.H. et al. (2001) Transgenic cystic fibrosis mice exhibit reduced early clearance of Pseudomonas aeruginosa from the respiratory tract. J. Immunol. 166, 7410–7418 24 van Heeckeren, A. et al. (1997) Excessive inflammatory response of cystic fibrosis mice to bronchopulmonary infection with Pseudomonas aeruginosa. J. Clin. Invest. 100, 2810–2815 25 Gosselin, D. et al. (1998) Impaired ability of Cftr knockout mice to control lung infection with Pseudomonas aeruginosa. Am. J. Respir. Crit. Care Med. 157, 1253–1262 26 Yu, H. et al. (2000) Innate lung defences and compromised Pseudomonas aeruginosa clearance in the malnourished mouse model of respiratory infections in cystic fibrosis. Infect. Immun. 68, 2142–2147 27 van Heeckeren, A. et al. (1998) Effects of bronchopulmonary inflammation induced by Pseudomonas aeruginosa on adenovirus mediated gene transfer to epithelial cells in mice. Gene Ther. 5, 345–351 28 Kelley,T.J. and Drumm, M.L. (1998) Inducible nitric oxide synthase expression is reduced in cystic fibrosis murine and human airway epithelial cells. J. Clin. Invest. 102, 1200–1207 29 Steagall, W.K. et al. (2000) Cystic fibrosis transmembrane conductance regulator-dependent regulation of epithelial inducible nitric oxide synthase expression. Am. J. Respir. Cell Mol. Biol. 22, 45–50 30 Wine, J. (1999) The genesis of cystic fibrosis lung disease. J. Clin. Invest. 103, 309–312
TRENDS in Genetics, Vol.17 No.10, October 2001
31 Cowley, E.A. et al. (2000) Airway surface liquid composition in mice. Am. J. Physiol. 278, L1213–L1220 32 McCray, P.B. et al. (1999) Efficient killing of inhaled bacteria in ∆F508 mice: role of airway surface liquid composition. Am. J. Physiol. 277, L183–L190 33 Zahm, J-M. et al. (2001) X-ray microanalysis of airway surface liquid collected in cystic fibrosis mice. Am. J. Physiol. Lung Cell. Mol. Physiol. 281, L309–L313 34 Jayaraman, S. et al. (2001) Airway surface liquid osmolality measured using fluorophore-encapsulated liposomes. J. Gen. Physiol. 117, 423–430 35 Pier, G.B. et al. (1997) Cystic fibrosis transmembrane conductance regulator is an epithelial cell receptor for clearance of Pseudomonas aeruginosa from the lung. Proc. Natl.Acad. Sci. U. S.A. 94, 12088–12093 36 Chroneos, Z.C. et al. (2000) Role of cystic fibrosis transmembrane conductance regulator in pulmonary clearance of Pseudomonas aeruginosa in vivo. J. Immunol. 165, 3941–3950 37 Gabriel, S.E. et al. (1994) Cystic fibrosis heterozygote resistance to cholera toxin in the cystic fibrosis mouse model. Science 266, 107–109 38 Pier, G.B. et al. (1998) Salmonella typhii uses CFTR to enter intestinal epithelial cells. Nature 393, 79–82 39 Snouwaert, J.N. et al. (1992) An animal model for cystic fibrosis made by gene targeting. Science 257, 1083–1088 40 Dorin, J.R. et al. (1992) Cystic fibrosis in the mouse by targeted insertional mutagenesis. Nature 359, 211–215 41 Ratcliff, R. et al. (1993) Production of a severe cystic-fibrosis mutation in mice by gene targeting. Nat. Genet. 4, 35–41 42 O’Neal, W.K. et al. (1993) A severe phenotype in mice with a duplication of exon 3 in the cystic fibrosis locus. Hum. Mol. Genet. 2, 1561–1569
A TRENDS Guide to Mouse Models of Human Diseases
|
Review
43 Hasty, P. et al. (1995) Severe phenotype in mice with termination mutation in exon 2 of cystic fibrosis gene. Somatic Cell Mol. Genet. 21, 177–187 44 Colledge, W.H. et al. (1995) Generation and characterisation of a delta-F508 cystic fibrosis mouse model. Nat. Genet. 10, 445–452 45 Zeiher, B.G. et al. (1995) A mouse model for the delta-F508 allele of cystic-fibrosis. J. Clin. Invest. 96, 2051–2064 46 van Doorninck, J.H. et al. (1995) A mouse model for the cystic fibrosis delta-F508 mutation. EMBO J. 14, 4403–4411 47 Clarke, L.L. et al. (1992) Defective epithelial chloride transport in a gene targeted mouse model of cystic fibrosis. Science 257, 1125–1128 48 Clarke, L.L. et al. (1994) Relationship of a non-CFTR mediated chloride conductance to organ level disease in cftr (−/−) mice. Proc. Natl.Acad. Sci. U. S.A. 91, 479–483 49 Smith, S.N. et al. (1995) Bioelectric characteristics of exon 10 insertional cystic fibrosis mouse: comparison with humans. Am. J. Physiol. 268, C297–C307 50 French, P.J. et al. (1996) A delta F508 mutation in mouse cystic fibrosis transmembrane conductance regulator results in a temperature-sensitive processing defect in vivo. J. Clin. Invest. 98, 1304–1312 51 Grubb, B.R. et al. (1994) Hyperabsorption of Na+ and raised Ca2+-mediated Cl− secretion in nasal epithelia of CF mice. Am. J. Physiol. 266, C1478–C1483 52 Wilschanski, M.A. et al. (1996) In vivo measurements of ion transport in long-living CF mice. Biochem. Biophys. Res. Commun. 219, 753–759 53 Thomas, G.R. et al. (2000) G551D cystic fibrosis mice exhibit abnormal regulation of inflammation in lungs and macrophages. J. Immunol. 164, 3870–3877
S37
A TRENDS Guide to Mouse Models of Human Diseases
|
Review
Mouse models of cystic fibrosis Donald J. Davidson and Mark Rolfe The development of mouse models for cystic fibrosis has provided the opportunity to dissect disease pathogenesis, correlate genotype and phenotype, study disease-modifying genes and develop novel therapeutics. This review discusses the successes and the challenges encountered in characterizing and optimizing these models. Cystic fibrosis (CF) is a common, lethal, autosomal recessive disorder caused by mutations in the CFTR gene, with the most common mutation (∆F508) occurring on ~70% of CF chromosomes. Dysfunction of the CFTR protein, which acts as an apically localized epithelial chloride ion channel, results in the classical manifestations of CF: salty sweat, pancreatic insufficiency, intestinal obstruction, male infertility, and severe pulmonary disease, with characteristic abnormalities in electrolyte transport. The most serious consequence is progressive and ultimately fatal inflammatory lung disease characterized by chronic microbial colonization and repeated acute exacerbations of pulmonary infection, with a distinctive spectrum of pathogens.These clinical manifestations show considerable variation between individuals because of an as yet incompletely understood combination of environmental factors, independently segregating disease-modifying genes, and differences between specific CFTR mutations. A combination of chest physiotherapy, aggressive antibiotic therapy and pancreatic enzyme supplementation constitutes the traditional mainstay of treatment for CF. Despite the success of these therapies, novel approaches are required, both to achieve further increases in life expectancy and to improve quality of life rather than simply to alleviate symptoms. As our understanding of the molecular and biological basis of CF becomes more comprehensive, so the goal of successful treatments becomes more accessible. Generating mouse models of cystic fibrosis The first mouse models of CF were created within three years of the isolation of the human CFTR gene. To date, 11 different mouse models of CF have been characterized (Table 1). These models can be categorized as those designed simply to disrupt Cftr expression and those that specifically model various human clinical mutations. The former can be further subdivided into those that used a replacement strategy to disrupt the Cftr gene, creating absolute nulls (with no normal CFTR production), and those that used insertion into the target gene without loss of genomic sequence, in which the mutants retain the
potential to produce low levels of normal Cftr mRNA by various mechanisms.These distinctions affect the phenotype of the models and must be recognized in any interpretation of the phenotypes observed.
In association with MKMD
Phenotypes of mouse models of cystic fibrosis Characterization of the different mouse models of CF has demonstrated most of the same primary phenotypes, including intestinal obstruction, reduced fertility and characteristic intestinal and airway electrophysiology, reproducing many of the manifestations of CF in humans. Important differences have been observed, however, both between different models (particularly in survival rates), and between mouse models of CF and human disease patterns (most clearly observed in pancreatic function and the murine pulmonary phenotype). The phenotypic variations between mouse models have been shown to relate to the specific mutation of Cftr generated, to environmental influences and to independently segregating modifier genes. Given rigorous evaluation, these differences provide the means to start dissecting the key components in disease pathogenesis; however, the need to clearly define the mouse model studied becomes obvious. In addition, the most robust phenotypes have provided in vivo models for the development and optimization of novel therapeutics.
http://research.bmn.com/mkmd
Survival and intestinal disease Intestinal pathology and the resultant mortality are the hallmarks of Cftr mutation in the mouse. Considerable variation in the specific pathology and the degree of severity has been reported between different models. In most cases, however, characterization of the mutant mice has revealed abnormal electrophysiological profiles, runting and failure to thrive, goblet cell hyperplasia, mucin accumulation, crypt dilation and intestinal obstruction (bearing similarity to meconium ileus, present in 10–15% of CF patients) with resultant perforation, peritonitis and death. Studies characterizing the electrophysiological profile have found broadly similar phenotypes in the different models (Table 2) with a significant decrease in the baseline potential difference (probably representing a decreased
0168-9525/01/$ – see front matter © 2001 Elsevier Science Ltd. All rights reserved. PII: S0168-9525(01)02452-0
Donald J. Davidson* University of British Columbia, BC Research Institute for Child and Family Health, Room 381, 950 West 28th Avenue, Vancouver, British Columbia, Canada V5Z 4H4. *e-mail: [email protected]
Mark Rolfe Medical Research Council Human Genetics Unit, Western General Hospital, Crewe Road, Edinburgh, UK EH4 2XU.
S29
Review
|
A TRENDS Guide to Mouse Models of Human Diseases
TRENDS in Genetics, Vol.17 No.10, October 2001
Table 1. Mouse models of cystic fibrosisa Mouse
Mutation
Cftr mRNA
Original strain
Cftrtm1Unc
(D) Exon 10 replacement
No wildtype (wt) mRNA detectable
C57Bl/6/129, BALB/c/129, B6D2/129
Refs 39
Cftrtm1Hgu
(D) Exon 10 insertional
10% of normal levels of wt mRNA
MF1/129
40
Cftrtm1Cam
(D) Exon 10 replacement
No wt mRNA detectable
MF1/129, C57Bl/6/129
41
Cftrtm1Hsc
(D) Exon 1 replacement
No wt mRNA detectable
CD1/129
3
Cftrtm1Bay
(D) Exon 3 insertional duplication
<2% of normal levels of wt mRNA
C57Bl/6/129
42
Cftrtm3Bay
(D) Exon 2 replacement
No wt mRNA detectable
C57Bl/6/129
43
Cftrtm2Cam
(C) ∆F508 Exon 10 replacement
Mutant mRNA 30% of normal expression levels
C57Bl/6/129
44
Cftrtm1Kth
(C) ∆F508 Exon 10 replacement
Decrease in mutant mRNA levels in intestinal tract
C57Bl/6/129
45
Cftrtm1Eur
(C) ∆F508 Exon 10 insertional 'hit and run’
Mutant mRNA expression at normal levels
FVB /129
46
Cftrtm1G551D
(C) G551D Exon 11 replacement
Mutant mRNA 53% of normal expression levels
CD1/129
6
Cftrtm2Hgu
(C) G480C Exon 10 insertional 'hit and run'
Mutant mRNA expression at normal levels
C57Bl/6/129
b
a
Mouse models of cystic fibrosis (CF) can be categorized as those that simply disrupt Cftr expression (D), and those that model specific clinical mutations (C). b Dickinson, P. et al., Generation of a CF mutant mouse possessing the G480C mutation. 22nd European Cystic Fibrosis Conference, 13–19 June 1998, Berlin.
rate of unstimulated Cl− secretion) and a complete absence, or a significant decrease, in cAMP-stimulated Cl− secretion (indicative of the loss of CFTR function). These profiles can be used to distinguish unequivocally mutants from their wild-type (wt) littermates and closely model the electrophysiological phenotype in CF humans. The survival rates of the different mouse models generally reflect the severity of intestinal pathology and vary from <5% in Cftrtm1Unc/Cftrtm1Unc nulls, and similar rates in the majority of models, to ~90% in Cftrtm1Hgu/Cftrtm1Hgu mice, and normal survival in Cftrtm2Hgu/Cftrtm2Hgu and Cftrtm1Eur/Cftrtm1Eur mice (Table 3). These dramatic differences appear to be the result of mutation-specific effects, independently segregating modifier genes and environmental influences. A low-level production of ~10% of normal CFTR has been proposed to be the explanation for the significantly greater survival rate in the residual function Cftrtm1Hgu/Cftrtm1Hgu mice1. In addition, in studies using mutant allele crosses, compound heterozygote mice have indicated that small levels of normal CFTR activity can have dramatic effects on survival2. These important studies suggest that gene therapy and other replacement or augmentation therapeutic strategies that are able to provide even modest restoration of function could provide significant benefits in CF individuals.The reason for the mild nature of the intestinal disease in the Cftrtm2Hgu/Cftrtm2Hgu and Cftrtm1Eur/Cftrtm1Eur mice is less clear. However, it could relate to the use of a double homologous recombination (‘hit and run’) strategy to generate these mice. As a result, both these mice
S30
express normal levels of a mutant CFTR protein, which could provide enough residual function in mice to change the phenotype from that observed in a ‘null’ animal or one expressing a low level of mutant protein. In studies using Cftrtm1Hsc/Cftrtm1Hsc mice bred to congenicity on different inbred backgrounds, the mortality has been shown to manifest at two distinct periods and to be partially determined by an independently segregating modifier locus3. These studies led to the identification of a genetic modifier locus for meconium ileus in humans4. This powerful technique to find disease-modifying genes is made possible in mice by the use of lines of genetically identical inbred animals. An important benefit of murine models of disease, this has the potential to provide great insights into the variability of disease manifestation and reveal possible targets for novel therapeutic approaches. Finally, survival has been shown to be influenced both by diet and housing conditions.The use of a liquid diet has been found to prolong the lifespan of Cftrtm1Unc/Cftrtm1Unc mice5, whereas the mortality observed in Cftrtm1G551D/ Cftrtm1G551D mice has been shown to be partially dependent upon the sterility of their housing conditions6. Analyses of these variable phenotypes has provided insights into the role of CFTR in the intestinal tract in vivo and demonstrated that this severe phenotype can be alleviated in mice by a low level of normal CFTR, normal levels of mutant CFTR or the presence of certain modifier genes. Although rather more severe than in humans, the intestinal phenotype in these mouse models is sufficiently
TRENDS in Genetics, Vol.17 No.10, October 2001
A TRENDS Guide to Mouse Models of Human Diseases
|
Review
Table 2. Intestinal electrophysiology in mouse models of cystic fibrosisa Mutation
Tissue
Baseline PD
cAMP-mediated Cl– response
Ca2+-related Cl– response
Refs
CF human
GI tract
↔ or ↑
↓
↓
Cftrtm1Unc
Jejunum
↓
↓ 100%
↓
47
Caecum
↓
↓ 100%
↓
48
Colon
↓
↓ 100%
↓
48
Jejunum
↓
↓ 65%
↓
49
Cftrtm1Hgu
Caecum
↓
↓ 65%
↓
49
Cftrtm1Cam
Caecum
↓
↓ 100%
n.r.
41
Cftrtm1Hsc
Rectum
n.r.
↓ 100%
↑
3
Ileumb
n.r.
↓ 100%
↑
3
Cftrtm1Bay
Ileum
↔
↓ 80%
n.r.
42
Cftrtm3Bay
Colon
n.r.
↓ 100%
↓
43
Cftrtm2Cam
Colon
↓
↓ 100%
↓
44
Cftrtm1Kth
Jejunum
↔
↓ 100%
n.r.
45
CftrtmEur
Ileum
↓
↓ 66%
↔
46
Caecum
n.r.
↓ 92%
n.r
50
Jejunum
↓
↓ 99%
n.r.
6
Caecum
↓
↓ 95%
n.r.
6
Caecum
↓
↔
↓
c
Cftrtm1G551D Cftrtm2Hgu aComparison
of the electrophysiological profiles of the intestinal epithelium in human cystic fibrosis (CF) and mouse models of CF, on the original background strain. Increased (↑), decreased (↓) or preserved (↔) potential difference (PD) in comparison with non-CF controls. Abbreviations: GI, gastrointestinal tract; n.r., not reported. bPatch-clamped, isolated ileal crypt cells. cDickinson, P. et al., Generation of a CF mutant mouse possessing the G480C mutation. 22nd European Cystic Fibrosis Conference, 13–19 June 1998, Berlin.
similar to suggest the same pathophysiological processes, validating their use as models for human disease and the relevance of these findings for novel therapy development. Pancreatic disease Pancreatic insufficiency is a prominent manifestation of CFTR dysfunction in humans, but has not been convincingly demonstrated in most mouse models of CF.This appears to be the result of low levels of expression of Cftr in the murine pancreas and the presence of an alternative fluid secretory pathway, which is activated by increases in intracellular calcium7.This indicates that other ion channels might be capable of compensating for the loss of CFTR and suggests novel therapeutic approaches in humans, namely to identify and utilize such pathways. One study of Cftrtm1Unc/Cftrtm1Unc mice weaned on a liquid diet to increase survival rates, demonstrated significant differences in pancreatic growth and specific enzyme activities8. Similar, although less severe, abnormalities in wt controls, however, suggested that the abnormalities were predominantly secondary to malnutrition. A further study using a liquid elemental diet reported luminal dilatation
and the accumulation of zymogen granules at the apical pole of the ductal epithelial cells in Cftrtm1Unc/Cftrtm1Unc mice. This phenotype has since been used, and corrected with oral administration of docosahexanoic acid (DHA), in a study of the role of dietary fatty acids in CF (Ref. 9). This study suggested that a primary defect in fatty acid metabolism might play a significant role in the pathogenesis of CF and indicated DHA as a novel therapy. However, the role of the liquid diet in this phenotype might yet prove to be significant. Lung disease Lung disease represents the primary concern in CF and the manifestation for which an animal model is likely to be the most valuable. Although a variety of pulmonary abnormalities have been reported, these are complex, mostly precipitated in response to exposure to pathogens, and do not fully replicate the pathogenesis of lung disease in CF individuals. Nevertheless, the development of lung phenotypes secondary to the loss of CFTR function in mice provides an important in vivo system to study the pathogenesis of CF lung disease.
S31
|
Review
A TRENDS Guide to Mouse Models of Human Diseases
TRENDS in Genetics, Vol.17 No.10, October 2001
Table 3. Survival in mouse models of cystic fibrosisa Mutation
Perinatal death
Survival to maturity
Body weight
∆F508 CF human
10% MI
20% DIOS
Failure to thrive
Cftrtm1Unc null
50% by day 7 40% death at weaning
<5% survival to maturity
10–50% reduction
39
Cftrtm1Hgu residual function
5% by day 7 2% death at weaning
90% survival to maturity
No reduction
40
Cftrtm1Cam null
80% by day 7 10% death at weaning
<5% survival to maturity
50% reduction
41
Cftrtm1Hsc null
55% by day 7 20% death at weaning
25% survival to maturity
Delayed
3
Cftrtm1Bay null
40% by day 7 10% death at weaning
n.r.
70% reduction
42
Cftrtm3Bay null
n.r.
40% survival at one month
Reduced
43
Cftrtm2Cam ∆F508
35% by day 16
<5% survival to maturity
n.r.
44
Cftrtm1Kth ∆F508
10% by day 7
40% survival to maturity
50% reduction
45
Cftrtm1Eur
∆F508
Refs
None
Normal
20% reduction
46
Cftrtm1G551D G551D
n.r.
67% survival at day 35 in SPF conditions 27% survival at day 35 in normal conditions
30–50% reduction
6
Cftrtm2Hgu G480C
None
Normal
No reduction
b
a
Comparisons between the survival rates of mouse models of cystic fibrosis (CF). Abbreviations: DIOS, distal intestinal obstruction syndrome; MI, meconium ileus; n.r., not reported; SPF, specific pathogen free. b Dickinson, P. et al., Generation of a CF mutant mouse possessing the G480C mutation. 22nd European Cystic Fibrosis Conference, 13–19 June 1998, Berlin.
Electrophysiological studies Electrophysiological analyses of nasal and tracheal airway epithelia (Tables 4 and 5) in mouse models of CF have been shown to differentiate clearly between mutants and wt littermates, even in the absence of gross pathology. These studies clearly demonstrate the basic ion-channel defect, with the nasal epithelium of mouse models of CF accurately replicating the human profile. Analysis of the trachea has, however, proved to be more complex, with important differences in both sodium- and chloride-ion transport observed between murine and human profiles. Indeed, it has been proposed that alternative chloride-ion pathways dominate over CFTR in the mouse trachea and might alleviate the effects of CFTR dysfunction10.The extent to which these observations are replicated in the lower airways remains unknown. Analyses of inbred mouse strains have revealed considerable variation and demonstrated that the ion transport properties in the murine airways are regulated by independently segregating modifier genes.Thus, the consequences of CFTR dysfunction in the trachea might vary considerably between different mouse models of CF.This raises the possibility of dissecting out the component parts of the electrophysiological response and establishing their relative contributions to disease pathogenesis. The ability to distinguish, unequivocally, mutant mice from wt littermates using electrophysiological profiles has
S32
been crucial in the use of mouse models of CF for testing the efficacy of novel therapies, particularly somatic gene therapy. The basis of the gene therapy strategy for CF is the prevention of disease development by direct replacement of CFTR gene function, irrespective of a complete understanding of disease pathogenesis. Gene correction strategies must therefore be demonstrated to be safe and effective so that intervention can be attempted in infants. Successful correction towards the wt pulmonary electrophysiological phenotype has provided the primary endpoint for analyses in mouse models of CF, using purified CFTR protein11 and both adenoviral-based12 and liposome-based13,14 gene therapy vectors. These studies were instrumental in the initiation of human gene therapy trials, provided results similar to those achieved in humans, and revealed many of the same obstacles to successful correction. These issues, such as low transfection efficiency, transient expression and effect of repeat administration, are now being addressed. Mouse models continue to prove an important resource in this development and optimization process and bridge the gap between cell-culture-based studies and human trials. Lung pathology The use of electrophysiological profiles has been of great importance in the development of novel therapies; however, disease state endpoints are also vital to evaluate the
TRENDS in Genetics, Vol.17 No.10, October 2001
A TRENDS Guide to Mouse Models of Human Diseases
|
Review
Table 4. Nasal electrophysiology in mouse models of cystic fibrosisa Mutation
Baseline PD
Amiloride response
cAMP-mediated Cl– response
Ca2+-related Cl– response
CF human
↑
↑
↓
↔
Cftrtm1Unc
↑
↑
↓ 100%
↑
51
Cftrtm1Hgu
↑
↑
↓ 70%
↔
49
Cftrtm1Cam
↑
n.r.
n.r.
n.r.
13
Cftrtm1Hsc
↑
↑
↓ 100%
↑
52
Cftrtm1Eur
↑
↑
Response to Cl– gradient
46
Cftrtm1Kth
↑
↑
↓ 100%
n.r.
45
Cftrtm1G551D
↑
↑
↓ 100%
↔
6
Refs
aComparison
of the electrophysiological profiles of the nasal epithelium in human cystic fibrosis (CF) and mouse models of CF, on the original mixed genetic background strains. Profiles are shown as increased (↑), decreased (↓) or preserved (↔) potential difference (PD) in comparison with non-CF controls. Amiloride inhibits the epithelial sodium channel EnaC. Abbreviation: n.r., not reported.
pathophysiological consequences. These have proved to be considerably more complex (Table 6). The first observations relating to abnormal lung pathology in mouse models of CF were made in outbred MF1/129 Cftrtm1Hgu/Cftrtm1Hgu mice. No gross lung disease was observed at birth, or in animals born and raised in isolator conditions15, but cytokine abnormalities were observed in mutant mice maintained in standard animal facilities16.These observations and the studies that followed suggest that an abnormal lung phenotype might not manifest without exposure to pathogens. In response to exposure to aerosolized clinical isolates of Staphylococcus aureus and Burkholderia cepacia, Cftrtm1Hgu/Cftrtm1Hgu mice demonstrated significantly impaired airway clearance and the
development of significantly more severe, pathogen-specific, lung pathology15.These observations in response to clinically relevant B. cepacia infection have recently been repeated and elaborated upon using Cftrtm1Unc mice maintained on a liquid diet17. This recent model also demonstrates an increased influx but suboptimal activation of inflammatory cells in the lungs of the mutant mice. In earlier studies, however, abnormal lung pathology was not observed in Cftrtm1Unc mice following exposure to S. aureus (Ref. 18). It is probable that mutation-specific effects, differences in the mouse-strain backgrounds (particularly variations in the role of alternative airway epithelial ion-channels) and varied methods of bacterial exposure influence these contrasting observations.
Table 5. Tracheal electrophysiology in mouse models of cystic fibrosisa Amiloride response
cAMP-mediated Cl– response
Ca2+-related Cl– response
Mutation
Baseline PD
CF human
↑ or ↔
↑
↓
↔
Cftrtm1Unc
↔
↔
↔
↔
10
Cftrtm1Hgu
↓
↓
↓ 60%
↔
49
Cftrtm1Cam
↓
↓
↓ 75%
↔
13
Cftrtm1Bay
↔
n.r.
↓b 70%
n.r.
42
Cftrtm2Cam
↔
↔
↔c
Cftrtm1G551D
↔
↔
↓ 60%
to ↓ 60%
Refs
↑
44
↑
6
a
Comparison of the electrophysiological profiles of the tracheal epithelium in human cystic fibrosis (CF) and mouse models of CF on the original mixed genetic background strains. Increased (↑), decreased (↓) or preserved (↔) potential difference (PD) in comparison to non-CF controls. Amiloride inhibits the epithelial sodium channel ENaC. Abbreviation: n.r., not reported. b Greatest decrease observed in the youngest mice. c Studied in cultured foetal tracheal cells.
S33
Review
|
A TRENDS Guide to Mouse Models of Human Diseases
TRENDS in Genetics, Vol.17 No.10, October 2001
Table 6. Lung phenotypes in mouse models of cystic fibrosisa Mutation
Observation
Refs
Cftrtm1Hgu
↓ Pulmonary clearance of Staphylococcus aureus and Burkholderia cepacia
15
↑ Pulmonary pathology after repeated exposure to S. aureus and B. cepacia
15
↑ Bronchoalveolar lavage tumour necrosis factor α
16
↑ Inflammatory cells in tracheal lamina propria
19
↓ Tracheal mucociliary transport
19
↔ Pulmonary clearance of Pseudomonas aeruginosa
b
Altered submucosal gland distribution
20
Cftrtm1G551D
↑ S100 Ca
53
Cftrtm1Unc
↓ Pulmonary clearance and ↑ pulmonary pathology after repeated exposure to B. cepacia 17
Cftrtm1Kth
2+
binding protein and tumour necrosis factor α
↔ Pulmonary clearance of P. aeruginosa and S. aureus
18
↓ Airway mucociliary transport
21
↑ Mortality from P. aeruginosa-laden agar beads
25
Cytokine abnormalities
24
↓ iNOS expression
29
↔ Pulmonary clearance of P. aeruginosa
22
↓ Lung cell ingestion, ↑ lung burden with P. aeruginosa
23
↓ iNOS expression
28
aComparison
of the pulmonary pathology observed in different mouse models of cystic fibrosis (CF). Observations: ↑, increased; ↓, decreased; ↔, no change in comparison to non-CF controls. bLarbig, M. et al., Pseudomonas aeruginosa infection in cystic fibrosis: animal model of the transgenic cftrm1HGU mouse. European Cystic Fibrosis Conference, 13–19 June 1998, Berlin.
Further studies using Cftrtm1Hgu mice have demonstrated significantly impaired mucociliary transport of inert particles in vivo (Ref. 19) and an altered distribution of mucus and antimicrobial airway surface liquid (ASL)-secreting submucosal glands20 suggesting possible mechanisms for disease development, secondary to CFTR dysfunction. A subsequent study using embedded, cultured lung slices from mice homozygous for the Cftrtm1Unc mutation, partially backcrossed onto the C57Bl/6 background, also demonstrated impaired mucociliary transport21. Although these studies provide evidence of an abnormal response to bacterial burden, initial studies examining the effect of exposure to the classic lung pathogen Pseudomonas aeruginosa failed to demonstrate any abnormal lung phenotype in Cftrtm1Hgu/Cftrtm1Hgu mice (Larbig, M. et al., Pseudomonas aeruginosa infection in cystic fibrosis: animal model of the transgenic cftrm1HGU mouse. European Cystic Fibrosis conference, 13–19 June 1998, Berlin), Cftrtm1Unc/Cftrtm1Unc mice18 or Cftrtm1Kth/Cftrtm1Kth mice (carrying the ∆F508 mutation)22. These studies suggested that, despite abnormal responses to other bacteria, mouse models of CF might not display increased susceptibility to pulmonary infection with P. aeruginosa. This is obviously a significant contrast to CF lung disease in humans, the explanation for which remains elusive. However, a recent study examining the in vivo significance of epithelial cell internalization of P. aeruginosa
S34
quantified and localized these organisms within a few hours after delivery23. Under the conditions of this study, significant differences were observed between various mouse models of CF and controls, suggesting that basic defects in the host interaction with this organism might indeed exist in the mouse models, despite the elusive nature of convincing pathology. The study of P. aeruginosa in mouse models of CF is probably complicated by the phenotypic alteration that this organism undergoes over the course of chronic infection of the lungs of CF individuals. Although the initial infections in CF individuals are with planktonic strains, rapid deterioration of the CF lung usually occurs after transformation of P. aeruginosa into the mucoid form in the host. It is therefore unclear whether the most revealing studies will arise from trying to mimic this process by challenging mice with non-mucoid P. aeruginosa to evaluate predisposition to infection, or modeling colonization with mucoid strains in the absence of previous rounds of infection. In addition, the significance of previous infections with other organisms, and the consequent antibacterial chemotherapy received, is unclear with respect to priming the CF lung for P. aeruginosa infection. To date, studies using agar beads laden with P. aeruginosa to model colonization have been more successful than the alternative approach. Using this technique, significantly
TRENDS in Genetics, Vol.17 No.10, October 2001
decreased survival rates have been demonstrated in Cftrtm1Unc/Cftrtm1Unc mice24,25; however, the basis for this observation is unclear. Bacterial proliferation was demonstrated regardless of genotype, only one of the studies recovered significantly higher numbers of bacteria from the lungs of mutant mice, and no significant differences in lung histopathology were observed between mutant mice and non-CF littermates. Perhaps the most significant observations were the significantly higher levels of proinflammatory cytokines and the decreased levels of the anti-inflammatory cytokine IL-10 in Cftrtm1Unc/Cftrtm1Unc mice, despite an absence of significant differences in the bacterial lung burden or pulmonary histopathology24. The role of the liquid diet fed to the mice in these studies could, however, prove to be significant, with a malnourished mouse model reported to demonstrate some striking similarities26. Nevertheless, the agar bead model reveals interesting phenotypic differences between mouse models of CF and control animals, and could prove effective in the study of the host response to established infection. This model has since been used in the further development of adenoviral-mediated gene transfer27. A further phenotype described in mouse models of CF is that of decreased pulmonary levels of the inducible isoform of nitric oxide synthase (iNOS), which has been implicated in the development of CF lung disease.The expression of iNOS is significantly reduced in mixed background strain Cftrtm1Kth/Cftrtm1Kth mice, homozygous for the ∆F508 mutation28, and Cftrtm1Unc/Cftrtm1Unc mice29. Furthermore, in Cftrtm1Unc/Cftrtm1Unc mice expressing human CFTR cDNA in the intestinal tract but not the nose, iNOS expression is observed in the ileum but not in the nasal epithelium29. The exact mechanism by which CFTR affects the expression of iNOS remains to be determined. In conclusion, despite some significant similarities between CF lung disease in humans and mouse models of CF and promising recent developments17,23, significant differences are also evident under the experimental conditions described, and the suitability of these models as endpoints for therapeutic testing remains controversial. Testing hypotheses in mouse models of cystic fibrosis Despite certain differences between the pathology observed in mouse models of CF and human CF individuals, the former clearly demonstrate a range of abnormal phenotypes as a result of Cftr mutation. Consequently, mouse models of CF have been used to perform in vivo analysis of several hypotheses describing the pathogenesis of CF that were generated primarily on the basis of in vitro studies. Recent hypotheses have emphasized the role of CFTR in determining either the volume or the ionic concentration of the ASL lining in the lung epithelia, affecting mucociliary clearance or the antibacterial activity of salt-sensitive peptides30. Contrasting hypotheses were established on the basis of in vitro studies and limited analysis of human
A TRENDS Guide to Mouse Models of Human Diseases
|
Review
ASL. Several studies have circumvented some of the complicating factors by measuring the ionic composition of ASL in airways of various mouse models of CF, using techniques that could not be performed in human subjects31–34. Although the results of these studies suggest that Cftr mutation does not affect ASL salt concentration, they conflict over the in vivo tonicity. However, these studies used contrasting techniques on mouse models of CF with different mutations in Cftr, on different background strains. All these potential influences, and others (e.g. the role and distribution of submucosal glands) can be controlled for, and altered, in future studies.Thus, the crucial factors influencing data from mouse model studies should be more easily addressed in future studies and provide in vivo answers to the questions raised by these hypotheses. In addition, lines of transgenic mice with ‘knockout’ mutations in genes encoding antibacterial peptides have been established and are currently being characterized.The analysis of these mutant mice and the offspring intercrossed with mouse models of CF will help to determine the role of these antibacterial agents in the lung, particularly with reference to CF lung disease. Mouse models of CF have also been used to examine the hypothesis that CFTR is a receptor for P. aeruginosa complete lipopolysaccharide core, resulting in bacterial internalization into epithelial cells. This mechanism has been proposed to perform a protective role in the normal lung and to be compromised in CF (Ref. 35). A recent study examined the host interaction with P. aeruginosa just 4.5 hours after bacterial delivery23.This study demonstrated significantly less ingestion of P. aeruginosa by lung cells and significantly greater bacterial lung burdens in various mouse models of CF compared to wt controls. Confocal and scanning electron microscopic imaging was used to demonstrate association between airway epithelial cells and bacteria in the wt controls, suggesting a possible role for these cells in host defense. By contrast, a study using a range of transgenic mice expressing varying levels of human and murine CFTR found no direct correlation between the level of CFTR expression and the pulmonary clearance of P. aeruginosa, or the association of bacteria with epithelial cells in vivo (Ref. 36). The time points and the choice of strains evaluated in these studies appear to be crucial23, however, emphasizing the requirement for rigorous evaluation in optimization of these phenotypes and perhaps the subtlety of the underlying defect. Mouse models of CF have also been used to examine possible hypotheses for the high frequency of the mutant CF allele in human Caucasian populations.These studies have suggested two possible heterozygote advantage theories. In Cftrtm1Unc/Cftrtm1Unc mice, expressing no Cftr, intestinal fluid secretion was not observed in response to cholera toxin, and in heterozygote mice secretion was reduced to 50% of normal wt levels37. This suggests that individuals carrying a mutant CF allele might be more resistant to the potentially fatal diarrhea and dehydration induced by
S35
Review
|
A TRENDS Guide to Mouse Models of Human Diseases
Vibrio cholerae. A later study demonstrated that homozygous and heterozygous Cftrtm2Cam mice were protected from intestinal epithelial invasion by Salmonella typhii, lending credence to the idea that mutant CF alleles might confer a resistance to typhoid fever38. Thus, the ability to pursue invasive techniques in vivo and to control for genetic and environmental influences in future studies makes mouse models of CF a powerful system to help dissect the pathogenesis of CF lung disease and define many of the critical components of disease development. Future directions The challenges for future use of mouse models of CF initially lie in refining existing models to replicate human disease as accurately as possible and then in understanding the mechanisms that underlie the development of the mutant phenotypes observed only in mice. The phenotypic variability observed in mouse models of CF as a result of different strain backgrounds, specific mutations in Cftr, and environmental influences, is of great potential for the definition of genetic modifiers and to refine the mouse models of CF. These processes should define many of the key components involved in disease pathogenesis, establish the most suitable models for therapy testing with clear relevant clinical endpoints, and suggest novel therapeutic approaches for the treatment of human disease. The future availability of comprehensive gene expression arrays could prove particularly illuminating in studying the mechanisms underlying the phenotypic differences observed between different models, and between mouse models of CF and non-CF littermates, in response to different environmental stimuli. By adopting such an approach, mouse models of CF can provide valuable contributions to our understanding of this disease process and support efforts towards organ-based treatment for CF patients. Acknowledgements We gratefully acknowledge Julia R. Dorin, David J. Porteous, Eric W.F.W. Alton and Steve Smith. Donald Davidson is a Wellcome Trust fellow. Mark Rolfe is supported by the Medical Research Council, UK. References 1 Dorin, J.R. et al. (1994) Long term survival of the exon 10 insertional cystic fibrosis mutant mouse is a consequence of low level residual wild type Cftr gene expression. Mamm. Genome 5, 465–472 2 Dorin, J.R. et al. (1996) A demonstration using mouse models that successful gene therapy for cystic fibrosis requires only partial gene correction. Gene Ther. 3, 797–801 3 Rozmahel, R. et al. (1996) Modulation of disease severity in cystic fibrosis transmembrane conductance regulator deficient mice by a secondary genetic factor. Nat. Genet. 12, 280–287 4 Zielenski, J. et al. (1999) Detection of a cystic fibrosis modifier locus for meconium ileus on human chromosome 19q13. Nat. Genet. 22, 128–129 5 Kent, G. et al. (1996) Phenotypic abnormalities in long-term surviving cystic fibrosis mice. Pediatr. Res. 40, 233–241
S36
TRENDS in Genetics, Vol.17 No.10, October 2001
6 Delaney, S.J. et al. (1996) Cystic fibrosis mice carrying the missense mutation G551D replicate human genotype-phenotype correlations. EMBO J. 15, 955–963 7 Gray, M.A. et al. (1995) Chloride channels and cystic fibrosis of the pancreas. Biosci. Rep. 15, 531–541 8 Ip, W.F. et al. (1996) Exocrine pancreatic alteration in long-lived surviving cystic fibrosis mice. Pediatr. Res. 40, 242–249 9 Freedman, S.D. et al. (1999) A membrane lipid imbalance plays a role in the phenotypic expression of cystic fibrosis in cftr−/− mice. Proc. Natl.Acad. Sci. U. S.A. 96, 13995–14000 10 Grubb, B.R. et al. (1994) Anomalies in ion transport in CF mouse tracheal epithelium. Am. J. Physiol. 267, C293–C300 11 Ramjeesingh, M. et al. (1998) Assessment of the efficacy of an in vivo CFTR protein replacement therapy in CF mice. Hum. Gene Ther. 9, 521–528 12 Grubb, B.R. et al. (1994) Inefficient gene transfer by adenovirus vector to cystic fibrosis airway epithelia of mice and humans. Nature 371, 802–806 13 Hyde, S.C. et al. (1993) Correction of the ion transport defect in cystic fibrosis transgenic mice by gene therapy. Nature 362, 250–255 14 Alton, E.W.F.W. et al. (1993) Non-invasive liposome-mediated gene delivery can correct the ion transport defect in cystic fibrosis mutant mice. Nat. Genet. 5, 135–142 15 Davidson, D.J. et al. (1995) Lung disease in the cystic fibrosis mouse exposed to bacterial pathogens. Nat. Genet. 9, 351–357 16 Morrison, G. et al. (1997) Analysis of the pulmonary inflammatory phenotype in the CF mutant mouse. Pediatr. Pulmonol. (Suppl. 14), 317 17 Sajjan, U. et al. (2001) Enhanced susceptibility to pulmonary infection with Burkholdera cepacia in Cftr−/− mice. Infect. Immun. 69, 5138–5150 18 Cressman,V.L. et al. (1998) The relationship of chronic mucin secretion to airway disease in normal and CFTR-deficient mice. Am. J. Respir. Cell Mol. Biol. 19, 853–866 19 Zahm, J.M. et al. (1997) Early alterations in airway mucociliary clearance and inflammation of the lamina propria in CF mice. Am. J. Physiol. 41, C853–C859 20 Borthwick, D.W. et al. (1999) Murine submucosal glands are clonally derived and show a Cftr dependent distribution pattern. Am. J. Respir. Cell Mol. Biol. 20, 1181–1189 21 Cowley, E.A. et al. (1997) Mucociliary clearance in cystic fibrosis knockout mice infected with Pseudomonas aeruginosa. Eur. Respir. J. 10, 2312–2318 22 McCray, P.B. et al. (1999) Efficient killing of inhaled bacteria in ∆F508 mice: role of airway surface liquid composition. Am. J. Physiol. 277, L183–L190 23 Schroeder,T.H. et al. (2001) Transgenic cystic fibrosis mice exhibit reduced early clearance of Pseudomonas aeruginosa from the respiratory tract. J. Immunol. 166, 7410–7418 24 van Heeckeren, A. et al. (1997) Excessive inflammatory response of cystic fibrosis mice to bronchopulmonary infection with Pseudomonas aeruginosa. J. Clin. Invest. 100, 2810–2815 25 Gosselin, D. et al. (1998) Impaired ability of Cftr knockout mice to control lung infection with Pseudomonas aeruginosa. Am. J. Respir. Crit. Care Med. 157, 1253–1262 26 Yu, H. et al. (2000) Innate lung defences and compromised Pseudomonas aeruginosa clearance in the malnourished mouse model of respiratory infections in cystic fibrosis. Infect. Immun. 68, 2142–2147 27 van Heeckeren, A. et al. (1998) Effects of bronchopulmonary inflammation induced by Pseudomonas aeruginosa on adenovirus mediated gene transfer to epithelial cells in mice. Gene Ther. 5, 345–351 28 Kelley,T.J. and Drumm, M.L. (1998) Inducible nitric oxide synthase expression is reduced in cystic fibrosis murine and human airway epithelial cells. J. Clin. Invest. 102, 1200–1207 29 Steagall, W.K. et al. (2000) Cystic fibrosis transmembrane conductance regulator-dependent regulation of epithelial inducible nitric oxide synthase expression. Am. J. Respir. Cell Mol. Biol. 22, 45–50 30 Wine, J. (1999) The genesis of cystic fibrosis lung disease. J. Clin. Invest. 103, 309–312
TRENDS in Genetics, Vol.17 No.10, October 2001
31 Cowley, E.A. et al. (2000) Airway surface liquid composition in mice. Am. J. Physiol. 278, L1213–L1220 32 McCray, P.B. et al. (1999) Efficient killing of inhaled bacteria in ∆F508 mice: role of airway surface liquid composition. Am. J. Physiol. 277, L183–L190 33 Zahm, J-M. et al. (2001) X-ray microanalysis of airway surface liquid collected in cystic fibrosis mice. Am. J. Physiol. Lung Cell. Mol. Physiol. 281, L309–L313 34 Jayaraman, S. et al. (2001) Airway surface liquid osmolality measured using fluorophore-encapsulated liposomes. J. Gen. Physiol. 117, 423–430 35 Pier, G.B. et al. (1997) Cystic fibrosis transmembrane conductance regulator is an epithelial cell receptor for clearance of Pseudomonas aeruginosa from the lung. Proc. Natl.Acad. Sci. U. S.A. 94, 12088–12093 36 Chroneos, Z.C. et al. (2000) Role of cystic fibrosis transmembrane conductance regulator in pulmonary clearance of Pseudomonas aeruginosa in vivo. J. Immunol. 165, 3941–3950 37 Gabriel, S.E. et al. (1994) Cystic fibrosis heterozygote resistance to cholera toxin in the cystic fibrosis mouse model. Science 266, 107–109 38 Pier, G.B. et al. (1998) Salmonella typhii uses CFTR to enter intestinal epithelial cells. Nature 393, 79–82 39 Snouwaert, J.N. et al. (1992) An animal model for cystic fibrosis made by gene targeting. Science 257, 1083–1088 40 Dorin, J.R. et al. (1992) Cystic fibrosis in the mouse by targeted insertional mutagenesis. Nature 359, 211–215 41 Ratcliff, R. et al. (1993) Production of a severe cystic-fibrosis mutation in mice by gene targeting. Nat. Genet. 4, 35–41 42 O’Neal, W.K. et al. (1993) A severe phenotype in mice with a duplication of exon 3 in the cystic fibrosis locus. Hum. Mol. Genet. 2, 1561–1569
A TRENDS Guide to Mouse Models of Human Diseases
|
Review
43 Hasty, P. et al. (1995) Severe phenotype in mice with termination mutation in exon 2 of cystic fibrosis gene. Somatic Cell Mol. Genet. 21, 177–187 44 Colledge, W.H. et al. (1995) Generation and characterisation of a delta-F508 cystic fibrosis mouse model. Nat. Genet. 10, 445–452 45 Zeiher, B.G. et al. (1995) A mouse model for the delta-F508 allele of cystic-fibrosis. J. Clin. Invest. 96, 2051–2064 46 van Doorninck, J.H. et al. (1995) A mouse model for the cystic fibrosis delta-F508 mutation. EMBO J. 14, 4403–4411 47 Clarke, L.L. et al. (1992) Defective epithelial chloride transport in a gene targeted mouse model of cystic fibrosis. Science 257, 1125–1128 48 Clarke, L.L. et al. (1994) Relationship of a non-CFTR mediated chloride conductance to organ level disease in cftr (−/−) mice. Proc. Natl.Acad. Sci. U. S.A. 91, 479–483 49 Smith, S.N. et al. (1995) Bioelectric characteristics of exon 10 insertional cystic fibrosis mouse: comparison with humans. Am. J. Physiol. 268, C297–C307 50 French, P.J. et al. (1996) A delta F508 mutation in mouse cystic fibrosis transmembrane conductance regulator results in a temperature-sensitive processing defect in vivo. J. Clin. Invest. 98, 1304–1312 51 Grubb, B.R. et al. (1994) Hyperabsorption of Na+ and raised Ca2+-mediated Cl− secretion in nasal epithelia of CF mice. Am. J. Physiol. 266, C1478–C1483 52 Wilschanski, M.A. et al. (1996) In vivo measurements of ion transport in long-living CF mice. Biochem. Biophys. Res. Commun. 219, 753–759 53 Thomas, G.R. et al. (2000) G551D cystic fibrosis mice exhibit abnormal regulation of inflammation in lungs and macrophages. J. Immunol. 164, 3870–3877
S37