Infection experiments were performed with two CF mouse models (a) B6.129P2(CF/3)-Cftr
, (b) Cftr
-Tg(FABPCFTR)1Jaw/J and their respective littermates. According to the nomenclature of Teichgräber et al. the mouse lines are called (a) Cftr
and (b) Cftr
, their non-CF littermates (a) B6 and (b) WT, respectively. In Cftr
mice the exon 10 of the Cftr gene had been disrupted by the insertion of the vector pMCIneoPolyA . Since those mice produced low levels of Cftr  but showed a mixed genetic background , from the original Cftr
mutant mouse, CF strain CF/3-Cftr
was established at the Institute of Laboratory Animal Science of the Hannover Medical School by brother-sister mating for more than 40 generations. Next, the congenic mouse inbred strain B6.129P2(CF/3)- Cftr
, which is used in this study, was generated by 40 backcross generations using CF/3-Cftr
as donor strain and C57BL/6J as recipient strain . Following the nomenclature of Teichgräber et al.  this strain is called Cftr
, syngenic C57BL6/J mice are called B6 and served as controls. Cftr
mice are regulary monitored for their congenic C57BL/6J status using 27 SNP markers and integrity of the mutant Cftr locus by intragenic microsatellite markers [24, 26, 27].
-Tg(FABPCFTR)1Jaw/J mice were obtained from the Jackson Laboratories. These mice, in the following called Cftr
, which are of a mixed genetic background consisting of C57BL/6, FVB/N and 129, are knock-outs for the murine Cftr gene, but express human CFTR in the gut under control of the FABP1 (fatty acid binding protein1) promoter, which prevents acute intestinal obstruction 1 [12, 18]. Mice were obtained homozygous and heterozygous for the Cftr
targeted mutation (tm/tm and tm/+) as well as homozygous and hemizygous for the FABP-hCFTR transgene (tg/tg and tg/0). Tm/+ tg/0 mice were used as parents to generate wildtype control mice (called WT). Genotyping was performed using the protocols provided by the JAX lab .
Mice were maintained at the Central Animal Facility of the Hannover Medical School, Carl-Neuberg-Str. 1, 30625 Hannover, Germany. They were held in groups of three to five animals animals in microisolator cages (910 cm2) with filter top lids and free access to sterilised standard laboratory chow (diet No. 1324, Altromin, Lippe, Germany) and autoclaved, acidulated water at 21 ± 2°C, 55 ± 5% humidity and a 10:14 light-dark-cycle. None of the CF mice showed gastrointestinal complications which would require a special diet. All mice were regularly monitored for infection by typical pathogens according to the FELASA recommendations . All procedures performed on mice were approved by the local district governments (AZ. 33.9-42502-04-08/1528) and carried out according to the ILAR guidelines for the care and use of laboratory animals . Experimental groups of mice were allocated by age of either younger than three (henceforth designated as young) or elder than six months (henceforth designated as old). This selection aimed to mimic the categories of Teichgräber et al.  that mice older than 4 months were susceptible to an infection with P. aeruginosa whereas the younger mice were more or less resistant.
Due to breeding limitations B6 mice showed a strong predominance of males and no 192 h value of old B6 exist.
Non-invasive head-out spirometry with 14 parameters was performed on conscious restrained mice . In brief, four mice were investigated in parallel by placing them in glass inserts with their heads protruding out through a set of membranes ensuring an airtight fit. Respiration caused air to flow through a pneumotachograph positioned above the thorax of the animals. The airflow was digitalized and analyzed with the Notocord Hem 188.8.131.52 software package (Notocord Systems SAS, Croissy Sur Seine, France).
Pseudomonas aeruginosa strain TBCF10839  was grown in Luria broth (LB) overnight at 37°C. The overnight culture was washed twice with the same volume of sterile PBS to remove cell detritus and secreted exopolysaccharides, then the optical density of the bacterial suspension was determined and the intended number of colony forming units (CFU) was extrapolated from a standard growth curve. Inocula with 6.0 × 105 CFU in 30 μl were prepared by dilution with sterile PBS. This infection dose is approximately one tenth of the LD50 of strain TBCF10839 for C57BL/6J mice and was able to produce a clinical infection without mortality.
Mice were infected intratracheally (i.t.) with 6.0 × 105 CFU of P. aeruginosa strain TBCF10839 under a light anaesthesia. For detailed description of the view-controlled intratracheal instillation see Munder et al. . To characterize the course of the bacterial infection, the body condition, weight, rectal temperature and lung function of the mice were evaluated as described previously . In brief the overall health of the animals was assessed by vocalisation, piloerection, posture, locomotion, breathing, curiosity, nasal secretion, grooming and dehydration. Dysfunctions in single parameters were assessed by zero, one or two points. The overall fitness of the mice was determined by adding the points resulting in the following score of the mouse body condition: unaffected (0-1); slightly affected (2-4); moderately affected (5-7); severely affected (8-10); moribund (≥ 11).
Non-invasive head-out spirometry. First, spirometric values of uninfected animals (B6, WT, Cftr
with age young: < 99 days and old: > 179 days) were averaged (median) from three independent measurements preformed on consecutive days. Prior measurements assured that the mice had adapted to the procedure. Lung function measurements of infection experiments were taken daily at the five days prior to inoculation and at time points 4, 6, 8, 10, 12, 18, 24, 48, 72, 96, 120, 144, 168, 192 hours post inoculation.
Forty-eight hours after challenge subgroups of mice were euthanized. Their left lungs were taken for the determination of bacterial counts and the right lungs were stained and examined histologically.
Pathohistology of the lungs
The right lungs were fixed via the trachea (4% paraformaldehyd), embedded in paraffin and stained with haematoxilin/eosin. One section was selected that showed aspects from all three lobes of the right lung. This slide was examined in twenty fields of view at a 100 fold magnification using a Zeiss Axiophot photomicroscope. Inflammation was assessed using a semi quantitative pathohistological score . Shortly, lung histological changes were scored on a scale from one to three points. Points were given separately for lung parenchyma, airways and lung vessels. The total score classified airway inflammation into the categories: almost not visible (0-5); slight (6-20) moderate (21-40); severe/profound inflammation (41-60). In the current study no more than a medium-grade inflammation was seen, appearing as a purulent alveolar pneumonia with peribronchiolar and perivascular inflammatory infiltrates.
Lung bacterial numbers
The left lungs of the euthanized mice were ligated, resected and homogenised. Aliquots were plated and bacterial numbers of whole organs were calculated. Previous experiments showed that the distribution of bacteria is approximately equal in left and right lungs after i.t. application (data not shown).
Each CF mouse model and its wild type controls were investigated separately by age group using non-parametric test statistics of SPSS 16 (Version 16.0.2, SPSS Inc, Chicago, USA). p-values (p < 0.05) with subsequent Bonferroni correction were calculated by 2-sided Monte Carlo simulations (100,000 simulations). Hereby groups were composed of equal numbers of mice (perfect match approach).