cystic fibrosis lung disease following infection with pseudomonas aeruginosa in cftr knockout mice...

Cystic fibrosis lung disease following infection

with Pseudomonas aeruginosa in Cftr knockout

mice using novel non-invasive direct pulmonary

infection technique

C Guilbault1, P Martin1, D Houle1, M-L Boghdady1, M-C Guiot2, D Marion1

and D Radzioch1

1Montreal General Hospital Research Institute; 2Montreal Neurological Institute, McGill UniversityHealth Center, Montreal, Quebec, Canada

Summary

To better understand the mechanism of lung infection with Pseudomonas aeruginosa(P. aeruginosa), many techniques have been developed in order to establish lung infection inrodents. A model of chronic lung infection, using tracheotomy to inoculate the bacteria, hasbeen extensively used in the cystic fibrosis (CF) mouse model of lung infection. The cysticfibrosis transmembrane channel (Cftr) knockout (KO) mice are smaller than normal miceand are more sensitive to housing and nutritional conditions, leading to small amounts ofanimals being available for experiments. Because of these characteristics, and because of theinvasiveness of the infection procedure which we, and others, have been using to mimic thelung infection, we sought to find an alternative way to study the inflammatory responseduring lung P. aeruginosa infection. The technique we describe here consists of the injectionof bacterial beads directly into the lungs through the mouth without the need of any trachealincisions. This technique of direct pulmonary delivery enables much faster infection of theanimals compared with the intratracheal technique previously used. The use of this lessinvasive technique allows the exclusion of the surgery-related inflammation. Our resultsshow that, using the direct pulmonary delivery technique, the KO mice were moresusceptible to P. aeruginosa lung infection compared with their wild-type (WT) controls, asshown by their increased weight loss, higher bacterial burden and more elevatedpolymorphonuclear (PMN) alveolar cell recruitment into the lungs. These differences areconsistent with the pathological profiles observed in CF patients infected with P. aeruginosa.Overall, this method simplifies the infection procedure in terms of its duration andinvasiveness, and improves the survival rate of the KO mice when compared with thepreviously used intratracheal procedure.

Keywords Mice; lung infection; Cftr knockout; Pseudomonas aeruginosa; agar beads

The cystic fibrosis (CF) disease is caused by adefect in the cystic fibrosis transmembrane

channel (CFTR) that functions as a chloridechannel regulated by cyclic AMP.Dysfunction of the CFTR protein resultsin salty sweats, pancreatic insufficiency,intestinal obstruction, male infertility andsevere pulmonary disease. Most of themorbidity and mortality in CF patients

Accepted 27 January 2005 r Laboratory Animals Ltd. Laboratory Animals (2005) 39, 336–352

Correspondence: Danuta Radzioch, McGill UniversityHealth Center, Montreal General Hospital Research Insti-tute, 1650 Cedar Avenue, Room L11-218, Montreal, QuebecCanada H3G 1A4. Email: [email protected]

result from pulmonary complications.Chronic infection of the lungs with mucoidstrains of Pseudomonas aeruginosa, whichtends to persist in most patients, results inan exaggerated neutrophilic inflammatoryresponse and in a dysregulated productionof cytokines: high levels of proinflammatorycytokines (interleukin (IL)-1, IL-6, IL-8,tumour necrosis factor (TNF)) and low levelsof anti-inflammatory IL-10 in bronchoal-veolar lavage fluids (BALF) (Johnson et al.1995, Morissette et al. 1996, Heeckeren et al.1997a, Stotland et al. 2000).

Since the discovery of the Cftr gene andbecause no natural animal model is known,a number of animal models have beendeveloped (for a review, see reference Nelsonet al. 2003). Animal models represent thebest surrogate for the complexities of thehuman system, providing that the resultsfrom experimental animal studies areextrapolated wisely. Most models developedwere murine models; a number ofinvestigators have generated Cftr geneknockout (KO) mice by targeted genedisruption (Snouwaert et al. 1992, O’Nealet al. 1993, Colledge et al. 1995, Davidsonet al. 1995, van Doorninck et al. 1995,Zeiher et al. 1995, Delaney et al. 1996,French et al. 1996). Although the generatedmice have most of the symptoms of CF,only very few of them display the CF lungphenotype (Nelson et al. 2003). However, theanalysis of susceptibility to infection withP. aeruginosa of various strains of micewas useful in the identification of genesother than Cftr that influence the severityof the CF disease. When the first KO micewere generated, their analysis clearlyshowed that mice are much more resistantto P. aeruginosa infection than human CFpatients. Unlike human CF patients,spontaneous colonization in the animalswith the typical CF pathogens, includingP. aeruginosa, has not been detected, perhapsdue to the maintenance of the KO mice inpathogen-free conditions. In the nextfew years, it became very apparent thatthe genetic background of the KO mice isvery important since many of the strains ofmice express alternative calcium-regulatedchloride channels (Gandhi et al. 1998).

Since the discovery of the Cftr gene, C57BL/6H/M-Cftr.ko mice described byKent and colleagues (Kent et al. 1997)represent a unique model of spontaneouslyoccurring CF lung disease. Our laboratoryparticipated in developing variousbackcrosses of KO mice includingthe C57BL/6H/M-Cftr.ko mice, and thestudies presented in this manuscript wereperformed using this unique spontaneouslung disease model.

To better understand the mechanism oflung infection with P. aeruginosa, manytechniques have been developed in order toestablish both acute and chronic lunginfection. Delivery of free P. aeruginosa(Jackson et al. 1967, Southern Jr et al. 1970,Schook et al. 1977, Dunn et al. 1985) directlyinto the nose or by aerosol techniques (Yuet al. 1998, McCray Jr et al. 1999, Chroneoset al. 2000) produced an acute infectionmodel. In order to create a model of chroniclung infection, Cash et al. (1979) developed arat model of lung infection usingP. aeruginosa embedded in agarose beads,which was later modified by Starke et al.(1987) for mouse models. The bacteria areinoculated intratracheally withimmobilizing agents such as agar, agarose orseaweed alginate. The entrapment ofP. aeruginosa in agar beads seems to slow thegrowth of the bacteria within the beadscompared with outside, resembling thebiofilm state (van Heeckeren & Schluchter2002), and this method has consequentlybeen extensively used in most of the CFmouse models of lung infection.

Adding to the complexity of mimickingP. aeruginosa lung infection in mousemodels of CF, the KO mice themselves arechallenging to work with. These mice aresmaller than normal mice and are moresensitive to housing and nutritionalconditions, leading to a small amount ofanimals being available for experiments.Because of these characteristics specific toKO mice, and also because of theinvasiveness of the infection procedurewhich we, and others, have been using toimprove persistence of the lung infection,we sought to find an alternative way to studythe lung inflammatory response during lung

Laboratory Animals (2005) 39

Non-invasive P. aeruginosa lung infection 337

P. aeruginosa infection. We saw a very cleardifference in susceptibility to lung infectionbetween KO and wild-type (WT) animalsusing a method of intratracheal instillation;however, we were wondering if the surgeryperformed during this procedure and theconsequently enhanced inflammationassociated with the healing process wouldinfluence the susceptibility of mice toinfection with P. aeruginosa. The originaltechnique detailed in this paper consists ofthe injection of bacterial beads directly intothe lungs through the mouth without theneed for any tracheal incisions. Moreover,this technique of direct pulmonary deliveryinfects the animals much faster comparedwith the intratracheal technique previouslyused. Since these KO mice are smaller, moresensitive and more fragile than WT controlmice, these ameliorations of the techniquebenefit them greatly and allow us to study alarger number of animals. We also observedmuch lower mortality rates, especially inKO mice. Finally, the use of a less invasivetechnique compared with the trachealsurgery also permits the exclusion of thesurgery-related inflammation while tryingto study the inflammatory response tobacterial infection.

Overall, we developed an infectiontechnique that allows the animals torecuperate faster from the infectionprocedure, creating a lung infectionthat resembles bacterial colonization withP. aeruginosa observed in humans. Thepurpose of this paper is to describe thedetails of this technique that, in ouropinion, is far more superior comparedwith previously available techniques.We also present a detailed comparisonof the inflammatory responses inducedby infection with P. aeruginosa using thismethod between KO and their littermatecontrols that is helpful in understandingthe basis of CF lung disease. Taken as awhole, the data presented in this manuscriptclearly show the usefulness of the des-cribed new method of lung infection tostudy the mechanisms involved in theregulation of inflammation induced byP. aeruginosa infection in our CF mousemodel.

Materials and methods

Mice

Age- and gender-matched C57BL/6 (WT)mice (n¼ 88 females, and n¼ 49 males)and C57BL/6-Cftr�/� (KO) inbred mice(n¼ 9 females, and n¼10 males) were keptmurine pathogen-, helicobacter- andparasite-free. They were housed (1–4animals/cage), bred and maintained in abarrier facility unit under specific pathogen-free conditions with a 12 h light/dark cycle,at 22721C and a relative humidity of5075%. The animals were kept inpolycarbonate micro-isolator cages (LabProducts, Maywood, NJ, USA) withsterile maple hardwood bedding (PWI,St-Hyacinthe, QC, Canada) or corn bedding(Anderson, Bestmonro, LA, USA) andmaintained in ventilated racks (LabProducts) with a cycle of 50 changes of freshhigh-efficiency particulate air filters (HEPA)filter air per hour in each cage. WT micewere fed with the NIH-31-modified mouse-irradiated diet (Harlan Teklad, Indianapolis,IN, USA), whereas KO mice were fed thePeptamen liquid diet (Nestle Canada,Brampton, ON, Canada), starting at 14 daysof age. The liquid diet, freshly made everymorning, was provided in 50mL centrifugetubes (Sarstedt, Montreal, QC). One weekprior to infection, the WT controls wereswitched from the solid diet to the Peptamenliquid diet. All mice had ad libitum access tosterile acidified water. Experimentalprocedures with the mice were conducted inaccordance with the Canadian Council onAnimal Care and with the approval of theAnimal Care Committee of the McGillUniversity Health Center, Montreal,Quebec, Canada.

P. aeruginosa

P. aeruginosa strain 508 was kindlyprovided by Dr Jacqueline Lagace (Universityof Montreal, Montreal, QC). This strainhas a mucoid appearance when grown onblood agar and was originally isolatedfrom the sputum of a CF patient atSte-Justine Hospital, Montreal, QC.Bacteria stocks are stored at �801Cuntil used.


338 C Guilbault et al.

Inoculum preparation

In order to establish a model of prolonged lunginfection, bacteria-impregnated agar beadswere freshly prepared the day before eachexperiment and stored at 41C overnight. A1mL aliquot of inoculum from the frozenstock was used to inoculate 250mL of 4%proteose peptone broth. The bacterial suspen-sion was then placed in a shaking incubatorovernight at 371C. A 100mL aliquot from thissuspension was used to inoculate 100mL ofbroth that was aliquoted in 15mL tubes(Sarstedt, Montreal, QC, Canada); 6mL pertube. Bacteria were grown for approximately3h in a shaking incubator at 371C, until theyreached a mid-log phase. Either eight or all ofthe 16 tubes containing log phase bacteriawere pooled together, depending on thedesired bacterial concentration of the beads.The bacteria were concentrated andresuspended in 5mL of Dulbecco’s phosphate-buffered saline (PBS; Invitrogen, Mississauga,ON, Canada). A 1 mL or 5mL aliquot of theconcentrated bacterial broth was added towarm (521C) 1.5% trypticase soy agar (TSA)(Difco, Detroit, MI, USA) (agarose beads freeof bacteria were prepared using PBS instead ofa bacterial suspension). This mixture wasquickly added to warm (521C) heavy mineraloil and stirred rapidly, first at room tempera-ture for 6min, followed by ice cooling withcontinuous stirring for 10min. The speed ofthe stirring was predetermined and remainedconsistent for every agarose–bacterialsuspension production. The oil–agar mixturewas centrifuged to sediment the beads. Theoil was removed and the beads were washedfour times with PBS. The size of the beadswas verified microscopically and only thosepreparations containing beads predominantly100–250mm in diameter were used asinoculum. The number of bacteria wasdetermined after homogenizing the bacteria-impregnated bead suspension. Inoculum wasprepared by diluting the beads suspension to0.5–20� 106 colony-forming units (CFU)/mL.

Description of the method of mouse lunginfection with P. aeruginosa instilled inagarose beads

Mice were anaesthetized with a combinationof ketamine (7.5mg/mL) and xylazine

(0.5mg/mL) administered intraperitoneallyat a dose of 20mL/kg of body weight. Weinitially tried a shorter acting anaesthetic,2,2,2 tribromoethanol (Avertin; SIGMA-Aldrich, Oakville, ON, Canada), but thisanaesthesia was not consistent. Once themouse was successfully anaesthetized (withno pedal or ocular reflex), the animal wasinstalled under binoculars (MicroscopeM650, Wild Leitz, Willowdale, ON, Canada)in the vertical position and was held on arestraining board by holding the animal byits upper incisor teeth, as shown in Figure1A. The tongue was then gently pulled to theside of the mouth (to depress the tongue inorder to see the vocal cords, Figure 1B). Acurved 26-G gavage needle was then insertedinto the mouth and guided through thepharynx to gently touch the vocal cords tosee the lumen of the trachea; the needle wasthen introduced into the trachea to reach thelung for the bilateral injection of theinoculum. Inoculums ranging from 30 to100 mL were used, but the 50mL volume wasthe preferred volume with regards to theconsistency of CFU injected. The injected50 mL volume did not lead to any distress inthe mice. After inoculation, the animalusually regained righting reflex within anhour. Many doses were tested using thistechnique, ranging from 0.5 to 20� 106

P. aeruginosa. A final dose of 1�106

P. aeruginosa was used for infection usingWT and KO mice. The weight of each mousewas recorded prior to and during the courseof infection.

Mice evaluation

After infection, the food was placed at thebottom of the cage, faecal and urine secre-tions were checked and removed everyday.Mice were monitored three times daily; themaximum weight loss allowed was 15%.Mice were sacrificed by CO2 overdose.

Bronchoalveolar lavage

Circulation was flushed by slow intracardiacinfusion of divalent cation-free Hank’sbalanced salt solution (HBSS; Invitrogen,Mississauga, ON, Canada). The trachea wascannulated with a 22-G intravenous catheter



placement unit (Critikon, GE MedicalSystems, Tampa, FL, USA) connected to two5mL syringes via a three-way stopcock witha rotating collar (Namic USA, Glens Falls,NY, USA). The alveoli of infected mice werewashed three times with 1.4mL of divalentcation-free HBSS. The volume of BALFrecovered was approximately 1.2mL.Alveolar cells were centrifuged and thesupernatant was used for CFU countdetermination before being stored at �201C,until assayed for cytokine concentrations.

Cells were resuspended in 0.5mL ofDulbecco’s modified Eagle’s medium(DMEM; Invitrogen) supplemented with10% fetal bovine serum (FBS; Hyclone,Logan, UT, USA), diluted in Turk’s solutionand counted using a haematocytometer. Theproportions of macrophages, lymphocytesand PMN were calculated after countingapproximately 300 alveolar cells on cytospinpreparations stained with Diff-Quick(American Scientific Products, McGaw Park,IL, USA).

Lung homogenates

Lungs from infected mice were harvestedand homogenized for 60 s at high speed(homogenizer PT10135 BrinkmannInstruments Co, Mississauga, ON, Canada)in 4mL of sterile PBS (Invitrogen). Serial10-fold dilutions of lung homogenates wereplated on Petri dishes containing TSA. Thenumber of CFU per lung was counted afterovernight incubation at 371C. For cytokinemeasurements, lung homogenates werecentrifuged at 1500 g at 41C for 10min; thesupernatants were then removed, aliquotedin new tubes and stored at �201C untilassayed for cytokine concentrations.

Lung histopathology

Lungs were removed from the mouse,inflated with 10% buffered formalin acetate(Fisher Scientific, Nepean, ON, Canada) andimmersed in that buffer for a minimum of36h. The lungs were then trimmed andembedded in paraffin. Paraffin sections weresliced 3mm thick using a Reichert–Jungmicrotome. Lung sections were cut atregular intervals to get sections at differentdepths of the lung. Three cuts were made foreach of the following standard stainingmethods: haematoxylin and eosin (HE),periodic-acid Schiff (PAS) and Masson’strichrome (MT). Each parameter of lungmorphology was observed under thesespecific stains and analysed using scalesspecific to each parameter. For hyperplasia ofthe epithelium and cell metaplasia, a scoreof 1 indicated no change; a score of 2represented 0–25% of the airway surface was



Figure 1 Direct pulmonary method of infection.Microscope and mouse holding system for lunginfection (A). View from the microscope of the vocalcords (B)

affected; a score of 3 represented 26–50% ofthe airway surface was affected; a score of 4meant that 51–75% of all airway surface wasaffected; and finally, a score of 5 indicatedthat 76–100% of the airway surface wascovered by these pathologies. Forinflammatory cell infiltration, basementmembrane thickening and fibrosis, the scalewas composed of scores 1, 2 and 3. A score of1 represented the absence of any of thesepathologies; a score of 2 indicated thepresence of either a few inflammatory cells,some basement membrane thickening orsome fibrosis (these last two seen thinlyaccumulated around the airways, showingthe airway wall about 2–5 times thicker thana control airway); a score of 3 was givenwhen the specimens showed the presence ofmany inflammatory cells in the bloodvessels around the airways or a significantamount of fibrosis and basement membranethickening around the airways (seen in thickdense layers around the airways, making theairway wall more than 5� thicker than anormal airway). Mucus accumulation wasanalysed separately for large airways(bronchi) and for smaller airways(bronchioles), because this phenomenonoccurred generally more often in largeairways than in smaller airways. No mucusaccumulation was expressed by a score of 1for small airways and a score of 4 for largeairways; some mucus accumulation(covering less than 50% of airway surface)was expressed by a score of 2 for smallairways and 5 for large airways; mucusaccumulation covering the airway surfacealmost completely was expressed by a scoreof 3 for small airways and 6 for large airways.

Cytokine measurements

The levels of 22 cytokines/chemokines (IL-1a, IL-1b, IL-2, IL-4, IL-5, IL-6, IL-7, IL-9, IL-10, IL-12 (p70), IL-13, IL-15, IL-17, interferon(IFN)-g, TNF, granulocyte–macrophagecolony-stimulating factor (GM-CSF),macrophage protein 1-alpha (MIP-1a),macrophage chemoattractant protein 1(MCP-1), cytokine-induced neutrophilchemoattractant (KC), Regulated onActivation Normal T Expressed and Secreted

(RANTES), interferon g-inducible protein 10(IP-10), granulocytes colony-stimulatingfactor (G-CSF)) were assessed in lunghomogenates prepared from infected animalswith the Mouse Cytokine/ChemokineLINCOplex kit (Linco Research Inc,St Charles, MO, USA) using LuminexTM

technology and assayed with theLuminex100ISTM system by Linco Research,Inc. The cytokine detection limit for thisassay was 3.2 pg/mL.

Statistical analyses

Data were analysed using Sigma Stat V3.01software (SPSS Inc, Chicago, IL, USA).Statistically significant differences betweenmeans and medians of studied groupswere evaluated using Student’s t-test andthe non-parametric Mann–Whitney U-test,respectively. One-way ANOVA andKruskal–Wallis ANOVA on ranks, combinedwith pair-wise multiple comparisonprocedures (Dunn’s method), were used toevaluate differences between multiplegroups. Significance was set at a two-tailedP value of p0.05.

Results

Method of infection

Mice are small animals compared with otherCF animal models like ferrets, cats, pigs,rabbits, sheep and rats. Therefore, we used amicroscope to properly inject the bacterialinoculums directly into the lung (Figure 1).In order to make sure that the injection wasinto the lungs exclusively, preliminaryexperiments were performed using a blue dyeto localize the site of inoculation. As shownin Figure 2, inoculation was well distributedthroughout the lungs. The injectionprocedure was performed numerous timesuntil fluidity in the mouse handling andperfect replicates were obtained.

P. aeruginosa dose determination for theestablishment of lung infection Since thisdirect pulmonary method of infection is lessinvasive than the intratracheal method, andbecause there is a susceptibility difference



between female and male mice toP. aeruginosa lung infection (Guilbaultet al. 2002), it was necessary to re-assess theappropriate dose of P. aeruginosa needed tocreate the pulmonary infection. Therefore,we evaluated numerous different doses,ranging from 0.5 to 80� 105 P. aeruginosaembedded in agarose beads, by comparingthe weight loss during infection, the bacter-ial burden and also the recruitment ofalveolar cells into the lungs (Figure 3) offemale and male WT mice (Table 1).

Bacterial burden in the lungs and survi-val The dose of P. aeruginosawe previouslyutilized in our CF mouse model is 1–2� 105

P. aeruginosa per mouse. Using this dose,almost no bacteria could be found in thelungs three days post-infection in males(Figure 3, panel B). Using our new method ofinfection, we observed almost completeclearance of bacteria from the lungs whendoses ranging from 2� 105 to 8� 105

P. aeruginosa were used, for both females(Figure 3, panel A; P¼ 0.181) and males(Figure 3, panel B; P¼ 0.293). At the 8� 105

P. aeruginosa dose and at all lower doses, allmice (100%) survived the infection proce-dure and recuperated from the intervention.

Given that the initial doses tried were notaffecting the survival rate of the mice andthat no major difference could be seen in thebacterial burden of the lungs, we consider-ably increased the dose of bacteria injectedinto the lung and applied doses ranging from0.8� 106 to 8� 106 P. aeruginosa per mouse.With the highest dose of P. aeruginosa tried,all mice died within a few hours followinginfection. Only 25% of the animals survivedwhen the dose of 5� 106 P. aeruginosa wasused (data not shown). When we infected themice with 2.5� 106 P. aeruginosa, 66% ofthe mice survived, but the majority of thesurviving animals lost more than 15% oftheir body weight and had to be sacrificed.When the mice were infected with 1� 106

P. aeruginosa, we observed a survival rateof 95.8%; therefore, this dose of bacteriaseemed to be the most appropriate for theanalysis of the inflammatory responses inthe lungs of the mice.

Alveolar cells in BALF samples The re-cruited inflammatory alveolar cells followedthe same trend seen in the CFU, as nostatistical difference was observed betweenthe doses ranging from 2� 105 to 8� 105

P. aeruginosa for females (Figure 3, panel C;



Figure 2 Site of injection into the lungs. A blue dye was injected into thelungs (volume between 30 and 50 mL) using the direct pulmonary injectiontechnique. Different views of the lungs are shown: Inside the mouse (A),frontal position (B) and dorsal position (C) of the lungs. After completenecropsy, no dye was found in the stomach of mice

P¼ 0.223) and males (Figure 3, panel D;P¼ 0.927). However, the number of inflam-matory cells in BALF samples from femalesincreased with the dose of P. aeruginosainjected into the lungs (Figure 3, panel C),

with a statistically significant increase(P¼ 0.010) found between the 8� 105

P. aeruginosa dose (1.4970.29 cells) andthe uninfected samples (0.1670.04 cells,data not shown).



A

C

E

B

D

F

Figure 3 Determination of the P. aeruginosa dose with direct pulmonary infectiontechnique. Female and male WT mice were infected with different doses of P. aeruginosaembedded in agarose beads and were examined at three days post-infection for bacterialburden in the lung (A and B), recruitment of inflammatory cells into the lung (C and D)and for weight loss during the course of infection (E and F). Increasing amounts of CFUand alveolar inflammatory cells were observed in the lungs correlating with P. aeruginosalung infection. Mice infected with P. aeruginosa lost weight during the first two days ofinfection, and then started regaining weight at day 3 post-infection, independently ofthe P. aeruginosa infection dose. Data are presented as the mean7SEM and pooled fromtwo different experiments done under the same conditions (females: 3–7 mice per group;males: 3–6 mice per group)

Weight monitoring during P. aeruginosa lunginfection Mice, both female and male,infected with P. aeruginosa lost between 7%and 12% of their body weight by day 2 post-infection and slowly started regainingweight by day 3 post-infection, independentof the dose of P. aeruginosa injected, whichranged from 2 to 8� 105 P. aeruginosa permouse (Figure 3, panels E and F). Also, nostatistical difference was observed in thetotal mean weight loss of the mice betweenthese three P. aeruginosa doses for all dayspost-infection analysed (day 1 post-infection,females P¼ 0.414 and males P¼ 0.182; day 2post-infection, females P¼ 0.097 and malesP¼ 0.390; and day 3 post-infection, femalesP¼ 0.171 and males P¼ 0.390). Mice injectedwith empty beads lost approximately 5% oftheir body weight the first day of infection,started gaining weight the second day afterinfection and regained it completely by day3 (data not shown).

In light of these results, the optimal dosechosen for the infection in our CF mousemodel was 1� 106 P. aeruginosa, as it created

a significant inflammatory response in thelungs of the mice without impairing theoverall wellbeing of the animal during theinfection and with the obtainment of areasonable survival rate.

Mouse model of CF

Histological evaluation of the lungs wasperformed on our CF mouse model in orderto identify the pathological state of theanimals before lung infection (Table 2). Asseen in Figure 4, the uninfected KO miceshowed increased hyperplasia of epithelialcells, basement membrane thickening(panels D and F), as well as enhancedinflammatory cell infiltration in the lungtissues (panel E) compared with the WTuninfected controls (panels A–C).

P. aeruginosa direct pulmonary infection inKO and WT mice

Next, using this newly developed andperfected technique, we infected KOmice and their WT controls with 1� 106

P. aeruginosa and evaluated theirinflammatory response.

Females displayed much highersusceptibility to P. aeruginosa lung infectionthan males in our mouse models (Guilbaultet al. 2002, and unpublished observations).Although we have used both genders todetermine the optimal and tolerable dosewith the direct pulmonary technique, whenstudying the CF lung disease in our CF



Table 2 Histopathological evaluation of lungs from non-infected KO and WT mice

Basement

Mucusaccumulation

Cellhyperplasia

Cellmetaplasia

membranethickening

Inflammatorycell infiltration

Smallairways

Largeairways Fibrosis

HE HE HE PAS PAS PAS MT HE

FemaleWT 1.8� 1.9 1.7 1.8w 1.4 4.7 1.3 1.4KO 2.3 1.4 1.5 2.5 1.3 4.7 1.3 1.3

MaleWT 1.0� 1.5 1.0 1.5w 1.5 5.0 N/D 1.5KO 2.8 1.5 1.7 2.1 1.3 4.4 1.5 1.5

HE=haematoxylin and eosin; PAS=periodic-acid Schiff; MT=Masson’s trichrome�Scores of the KO are significantly different compared with the WT for females (P=0.005) and males (P=0.041)wScores of the KO are significantly different compared with the WT for females (P=0.015) and males (P=0.035)

Table 1 Characteristics of mice used in directpulmonary infection procedures

Females Males

Age (weeks) 21.670.88 18.670.58Body weight

before infection (g)27.570.46 22.270.61

Values are means7SEM and represent the pool of animals from nineindependent experiments

mouse model, we predominantly used malesdue to a very high mortality rate observed infemales. Consequently, the figures includedin this section illustrate only resultsgenerated using male KO and littermatecontrols and the results from the females areonly described.

Weight monitoring during P. aeruginosa lunginfection The mean weight loss of KOmicewas shown to be higher than those seen inWT mice at two days (Pp0.001) and threedays (P¼ 0.001) post-infection for both males(Figure 5, panel A) and females (data notshown). A significant difference was alsoseen between the female and the male mice,for both WT and KO animals, as females lostmore weight than males (data not shown,day 2 post-infection Pp0.001 and day 3 post-infection P¼ 0.001). No difference was ob-served between any of the KO and WTgroups for both genders at day 1 post-infection (P¼ 0.070).

Bacterial burden in the lungs and survi-val The CFU counts were assessed inthe lung tissue homogenates of KO andWT males three days after infection.

A significantly higher number of bacteriawas found in the KO compared with the WTmales in the lung tissue homogenates (Fig-ure 5, panel B; P¼ 0.039). A difference wasalso observed between females and males,where females had between 30- and 300-foldhigher numbers of bacteria in the lungscompared with the male mice. Yet, becauseof the high variability seen in the females interms of CFU, no statistical significancecould be reached (data not shown, P¼ 0.086).

Alveolar inflammatory cells in BALF sam-ples The number of alveolar cells from theBALF was evaluated in KO and WT micethree days after infection with 1� 106

P. aeruginosa embedded in agar beads. Nodifference was found between the KO andWT for both male (Figure 5, panel C;P¼ 0.155) and female animals (data notshown, P¼ 0.700). Moreover, no differencewas observed between the genders for bothgenotypes in term of recruitment of cellsinto the lungs (data not shown, WT P¼ 0.683and KO P¼ 0.946). However, when wecompared the different types of inflamma-tory cells found in the lungs, we observedsignificant differences between the WT and



Figure 4 Histological examination of the lungs of uninfected WT and KO mice. RepresentativeHE-stained lung sections that were prepared from uninfected WT (A–C) and KO (D–F) mice.Sections from WT mice show a normal lung structure. Sections from KO mice show cell infiltrationin the bronchi and alveoli, and hyperplasia of the epithelium. Panels C and F illustrate a highermagnification of the airways of uninfected WT and KO mice showing normal and hyperplasia ofthe epithelium, respectively



.

.

.

.

A

B

D

C

Figure 5 Inflammatory response in WT and KO mice during lung infectionwith 1�106 P. aeruginosa embedded in agar beads. Using the direct pulmonaryinfection technique, WT (grey) and KO (white) male mice were evaluatedduring P. aeruginosa lung infection. Mean weight loss is represented as thepercentage of weight loss since the first day of infection (A). The CFU countswere assessed in the lung tissue homogenates three days after infection anddata are presented as mean7SEM (B). The number of alveolar cells from theBALF was evaluated three days after infection and data are presented asmean7SEM (C). The recruitment of inflammatory cells in the lungs of WT andKO mice was evaluated three days after infection: the numbers of alveolarmacrophages, lymphocytes and PMN were calculated on the basis of thedifferential cell counts of Diff-Quick cytospin preparations (D). These resultsillustrate two independent experiments (n¼3 per group for each experiment)performed under the same conditions

the KO mice in the mean percentages of thedifferent groups of cells (Figure 5, panel D).KO mice had higher PMN counts (P¼ 0.044),lower lymphocyte counts (P¼ 0.033) and aclear trend towards lower macrophagecounts as well, even though that lastdifference was not statistically significant(P¼ 0.107). In addition, WT female micerecruited more PMN and less lymphocytesin the lungs compared with the WT malemice, comparable to the levels observed inthe KO males (data not shown).

Cytokine expression in the lungs of P.aeruginosa-infected WTand KO mice Lunghomogenates from male mice were tested for22 different cytokines. Statistically higherlevels were observed in KO mice compared

with their WTcontrols for KC (P¼ 0.041), IL-9 (Pp0.001), GM-CSF (Pp0.001), IL-15(P¼ 0.008) and IL-7 (Pp0.001) (Figure 6). Allother cytokines measured were either notdetectable or not significantly different be-tween WT and KO mice at the timepointtested (Table 3).

Discussion

With many possible mutations in the Cftrgene, cystic fibrosis is a complex disease,affecting many organs with varyingintensities, frequently leading to death.What makes this disease difficult to study isthe fact that no known natural animal modelexists. Although many animal models havebeen used to study CF disease, the Cftr gene



Figure 6 Cytokines expression into the lungs of WT and KO infected mice. Lunghomogenates from male WT (solid) and KO (open) mice infected with 1�106 P.aeruginosa were tested for 22 different cytokines. Statistically significant differences(�) were observed in the levels of protein expression between the KO mice comparedwith their WT controls for KC (P¼0.041), IL-9 (Pp0.001), GM-CSF (Pp0.001), IL-15(P¼0.008) and IL-7 (Pp0.001). Data are presented as individual values and thehorizontal line represents the median. These results illustrate two independentexperiments performed under the same conditions

KO mouse model seems to be the mostfrequently used when researching thedifferent aspects of the disease and trying tobetter understand the CF lung disease.Various original techniques (acute andchronic models) using P. aeruginosa andBurkholderia cepacia (Urban et al. 2004)were developed for mice to mimic theinfection of the lungs of CF patients.However, even though rodents are easier towork with compared with other animalmodels, they are much more resistant to allkinds of microbes and microorganisms whencompared with humans. It is, therefore,quite challenging to establish conditionsallowing the lung bacterial colonization ofmice. Nevertheless, many techniques weredeveloped and are routinely used to studythe progress of CF lung disease. Two of thesetechniques involve the inhalation by mice ofthe bacteria following the aerosolization orthe inoculation of bacteria through the nose(Sokol et al. 2003). These models of infectionare primarily used to study the acuteinflammatory response in the lungs, sincethe animals clear the bacteria very rapidly.Several investigators, including us, haveused a technique of intratracheal infection of

the lungs with P. aeruginosa bacteriaembedded in agar beads (Gosselin et al. 1995,1998, Heeckeren et al. 1997b, Kent et al.1997, Van Heeckeren et al. 2000, Guilbaultet al. 2002). This method of infectioninvolves the inoculation of the bacteria intothe lungs via the cut-opened trachea. It alsorequires that the bacteria are enmeshed intoagarose or agar, allowing the bacteria to staylonger in the lung environment, mimicking,at least to some extent, the chronicity of thelung infection observed in CF afflictedpatients. This method has proven to beuseful in the study of bronchopulmonarychronic infection. However, a more ‘natural’colonization technique was used in a ratmodel of CF and other lung studies, invol-ving direct inoculation of the microorganisminto the lungs via the mouth without atemporary tracheotomy (Weksler et al.1994). This method of delivering the bacteriadirectly into the lungs without surgicalincision of the trachea was very appealing tous since it led to much less physical andinflammatory stress in the animals.

The difficulty lies in the fact that it ismuch more challenging to adapt thisprocedure for a mouse model due to the



Table 3 Cytokine levels in lung homogenates of P. aeruginosa infected KO andWTmice 3 days post-infection

WT KO

Quartiles Quartiles

Cytokines pg/mL 25% 75% pg/mL 25% 75%

MIP-1a� 191.7 88.7 216.1 255.7 88.2 393.9MCP� 108.9 64.2 125.6 137.4 61.3 187.4RANTES� 19.2 12.6 21.5 10.6 9.3 14.6IFNg� 9.1 6.3 10.7 10.1 2.5 10.4IL-1b� 29.9 15.9 42.8 24.4 9.7 45.9G-CSF� 155.1 70.2 184.1 270.0 117.8 433.2IP-10� 189.7 112.9 216.4 159.5 81.7 190.5IL-1a� 975.5 397.9 1403.5 1194.4 375.5 1945.8IL-6� 60.6 27.5 138.9 123.9 47.9 188.5IL-10� 0.0 0.0 0.0 19.3 4.8 21.2TNFa� 7.4 1.9 8.9 6.2 1.5 8.7IL-2 ND NDIL-4 ND NDIL-5 ND NDIL-12 ND NDIL-13 ND ND

Values are medians of 3–6 animals with quartilesND¼ levels of cytokine is below the level of detection of the assay (o3.2 pg/ml)�Medians are not significantly different at day 3 post-infection (P values between 0.200 and 1.000) in the KO compared with the WT mice

difference in size between a mouse tracheaand a rat trachea. Another study, workingwith a model of immunocompromised host,has used a similar methodology usingKlebsiella pneumoniae in B6D2F/J mice(Keller et al. 2003). The CF mouse modelproves to be even more challenging since theKO mice are much smaller than normalmice, and consequently they have smallertracheas than WT mice. KO mice are alsomore sensitive to manipulations than WTmice. Therefore, lesser amount ofmanipulations of the animals mightpotentially lead to a smaller mortality rate,and consequently a greater number of KOmice available for analysis. We then decidedto adapt this method so it could be used inour CF mouse model of lung infection withP. aeruginosa.

As described in this paper, theanaesthetized mouse was installedcomfortably in an upright position with itsback leaning against a solid metal holder.The mouse was maintained in that positionby a modified paper clip attached with a ropeto the back holder. The paper clip alsoinsured that the mouth was kept openduring the procedure with a clear view of thevocal cords. The use of a microscope allowedfor faster intervention since it gave precisionto the operator. The overall time of theprocedure was less than 2min/animal, fromthe moment of installation of the mouse onthe holder to the post-infection recoverystation. The mouse’s breathing rhythm wasuninterrupted. Interestingly, this techniqueallowed the use of a higher dose of bacteriabefore significant mortality was observedcompared with the surgical instillationmethod previously used in our C57BL/6mouse colony. At 1–2� 105 dose ofP. aeruginosa (using the direct pulmonarytechnique), all animals survived and hadcleared the bacteria by the third to the fifthday post-infection. At that dose of 1� 106

P. aeruginosa per mouse, we were able toestablish a long-lasting infection in the lungsmimicking the inflammatory response in CFlung disease. Doses higher than 1� 106

P. aeruginosa per mouse lead to eitherexcessive weight loss, choking or drasticmortality rates.

Once the P. aeruginosa infection dose wasestablished, we were able to evaluate thedifferences in lung pathology between KOmice and their WT controls. The primarygoal was to establish a long-lasting lunginfection that would reproduce, at least insome aspects, CF lung disease. While theinfection method does not involve anysurgical procedure, it still induces between3% and 14% weight loss during the firstthree days post-infection. WT animals lostsignificantly less weight during infectioncompared with KO mice (Figure 5). Whenusing the direct pulmonary technique ofinfection, we also found significantly higherbacterial burden in KO mice compared withWT mice. Although the amount of alveolarcells recruited to the lungs was not differentbetween WT and KO mice, we observed aclear imbalance in the types of inflammatorycells recruited, with relatively more PMNand a trend towards less macrophage countsin the lungs of the KO mice compared withtheir WT controls. Using the directpulmonary method of lung infection, wewere able to establish a similar immuneresponse to P. aeruginosa, as we observed inCF patients infected with P. aeruginosa. TheCF patients also displayed imbalanced influxof PMNs compared with macrophages in thelung, compared with non-CF patients.

Our mouse model of CF is one of a fewmodels known to develop lung disease evenprior to their infection with bacteria.Therefore, it was very interesting to evaluatethe differences in the cytokine profilebetween the KO and their WT controlsduring infection using the direct pulmonarytechnique. Cytokines, including KC, amurine homologue of IL-8 (Bonfield et al.1995, Dosanjh et al. 1998, Van Heeckerenet al. 2000, McMorran et al. 2001), IL-9, amediator of airway hyper-responsiveness andmucus overproduction (Hauber et al. 2003,2004), as well as GM-CSF (Sadikot et al.2004), known to be differently expressed inthe lungs of CF patients, were also found tobe expressed at a higher level in our KOmice. We have only analysed the levels ofcytokines secreted at one timepoint (day 3post-infection) due to the limited amount ofKO mice available. It is possible and very



likely, based on our previous publishedobservations (Gosselin et al. 1998), thatthree days post-infection timepoint mightnot be also appropriate for monitoring thelevels of IL-1b and IL-10, which were eithernot detectable or not significantly differentbetween WT and KO mice at the timepointtested (Table 3). Interestingly, we found a fewcytokines less studied in the context of CF,to be significantly different between KO andWT infected mice. One of them was IL-7, acytokine known to play an essentialfunction in T-cell development. Another wasIL-15, shown to stimulate the growth ofnatural killer cells, activated peripheralblood T lymphocytes (Giri et al. 1994, 1995,Grabstein et al. 1994), tumour infiltratinglymphocytes (TILs) (Lewko et al. 1995), andB cells (Armitage et al. 1995), was alsodifferently expressed in the lungs of infectedKO as compared with controls.

Overall, we describe in this paper amethod allowing us to establish long-lasting lung infection, creating differentinflammatory response in WT and KOmice infected with P. aeruginosa. TheKO mice were more susceptible toP. aeruginosa lung infection comparedwith their WT controls, as shown by theirincreased weight loss, higher bacterialburden and more elevated PMN alveolarcell recruitment into the lungs. Thesedifferences are consistent with thepathological profiles observed in the CFpatient infected with P. aeruginosa. Thismethod simplifies the infection procedurein terms of its duration, invasiveness,and improves the survival rate of the CFanimals when compared with the previouslyused intratracheal procedure. We hopethat the use of this method will stimulateeven more research using mice as a modelof lung infection, to better understand themolecular basis of CF lung disease andeventually help to develop successfultherapies applicable to CF patients infectedwith P. aeruginosa.

Acknowledgements This work was supportedby Canadian Cystic Fibrosis Foundation grant. DRwas also supported by the Chercheurs Nationauxscholarship from FRSQ-FCAR. We would like to

thank Claude Lachance for critical review of themanuscript.

References

Armitage RJ, Macduff BM, Eisenman J, Paxton R,Grabstein KH (1995) IL-15 has stimulatory activityfor the induction of B cell proliferation anddifferentiation. Journal of Immunology 154,483–90

Bonfield TL, Panuska JR, Konstan MW, et al. (1995)Inflammatory cytokines in cystic fibrosis lungs.American Journal of Respiratory and Critical Care

Medicine 152, 2111–18 [Erratum (1996): AmericanJournal of Respiratory and Critical Care Medicine

154 (4 Pt 1), following 1217]Cash HA, Woods DE, McCullough B, Johanson WG Jr,

Bass JA (1979) A rat model of chronic respiratoryinfection with Pseudomonas aeruginosa. Ameri-

can Journal of Respiratory Disease 119, 453–9Chroneos ZC, Wert SE, Livingston JL, Hassett DJ,

Whitsett JA (2000) Role of cystic fibrosis trans-membrane conductance regulator in pulmonaryclearance of Pseudomonas aeruginosa in vivo.Journal of Immunology 165, 3941–50

Colledge WH, Abella BS, Southern KW, et al. (1995)Generation and characterization of a delta F508cystic fibrosis mouse model. Nature Genetics 10,445–52

Davidson DJ, Dorin JR, McLachlan G, et al. (1995)Lung disease in the cystic fibrosis mouse exposedto bacterial pathogens. Nature Genetics 9, 351–7

Delaney SJ, Alton EW, Smith SN, et al. (1996) Cysticfibrosis mice carrying the missense mutationG551D replicate human genotype–phenotype cor-relations. EMBO Journal 15, 955–63

Dosanjh AK, Elashoff D, Robbins RC (1998) Thebronchoalveolar lavage fluid of cystic fibrosis lungtransplant recipients demonstrates increased in-terleukin-8 and elastase and decreased IL-10.Journal of Interferon Cytokine Research 18,851–4

Dunn MM, Toews GB, Hart D, Pierce AK (1985)The effects of systemic immunization ofpulmonary clearance of Pseudomonas aeruginosa.American Review of Respiratory Disease 131,426–31

French PJ, van Doorninck JH, Peters RH, et al. (1996)A delta F508 mutation in mouse cystic fibrosistransmembrane conductance regulator results in atemperature-sensitive processing defect in vivo.Journal of Clinical Investigation 98, 1304–12

Gandhi R, Elble RC, Gruber AD, et al. (1998)Molecular and functional characterization of acalcium-sensitive chloride channel from mouselung. Journal of Biological Chemistry 273,32096–101

Giri JG, Ahdieh M, Eisenman J, et al. (1994) Utiliza-tion of the beta and gamma chains of the IL-2



receptor by the novel cytokine IL-15. EMBO

Journal 13, 2822–30Giri JG, Kumaki S, Ahdieh M, et al. (1995) Identifi-

cation and cloning of a novel IL-15 binding proteinthat is structurally related to the alpha chain of theIL-2 receptor. EMBO Journal 14, 3654–63

Gosselin D, DeSanctis J, Boule M, Skamene E,Matouk C, Radzioch D (1995) Role of tumornecrosis factor alpha in innate resistance to mousepulmonary infection with Pseudomonas aerugino-sa. Infection and Immunity 63, 3272–8

Gosselin D, Stevenson MM, Cowley EA, et al. (1998)Impaired ability of Cftr knockout mice to controllung infection with Pseudomonas aeruginosa.American Journal of Respiratory and Critical Care

Medicine 157, 1253–62Grabstein KH, Eisenman J, Shanebeck K, et al. (1994)

Cloning of a T cell growth factor that interactswith the beta chain of the interleukin-2 receptor.Science 264, 965–8

Guilbault C, Stotland P, Lachance C, et al. (2002)Influence of gender and interleukin-10 deficiencyon the inflammatory response during lung infec-tion with Pseudomonas aeruginosa in mice.Immunology 107, 297–305

Hauber HP, Manoukian JJ, Nguyen LH, et al. (2003)Increased expression of interleukin-9, interleukin-9receptor, and the calcium-activated chloride chan-nel hCLCA1 in the upper airways of patients withcystic fibrosis. Laryngoscope 113, 1037–42

Hauber HP, Tsicopoulos A, Wallaert B, et al. (2004)Expression of HCLCA1 in cystic fibrosis lungs isassociated with mucus overproduction. EuropeanRespiratory Journal 23, 846–50

Heeckeren A, Walenga R, Konstan MW, Bonfield T,Davis PB, Ferkol T (1997a) Excessive inflammatoryresponse of cystic fibrosis mice to bronchopul-monary infection with Pseudomonas aeruginosa.Journal of Clinical Investigation 100, 2810–15

Heeckeren A, Walenga R, Konstan MW, Bonfield T,Davis PB, Ferkol T (1997b) Excessive inflammatoryresponse of cystic fibrosis mice to bronchopul-monary infection with Pseudomonas aeruginosa.Journal of Clinical Investigation 100, 2810–15

Jackson AE, Southern PM, Pierce AK, Fallis BD,Sanford JP (1967) Pulmonary clearance of Gram-negative bacilli. Journal of Laboratory and ClinicalMedicine 69, 833–41

Johnson LG, Boyles SE, Wilson J, Boucher RC (1995)Normalization of raised sodium absorption andraised calcium-mediated chloride secretion byadenovirus-mediated expression of cystic fibrosistransmembrane conductance regulator inprimary human cystic fibrosis airway epithelialcells. Journal of Clinical Investigation 95,1377–1382

Keller CE, Elliott TB, Bentzel DE, Mog SR, ShoemakerMO, Knudson GB (2003) Susceptibility of irradiatedB6D2F1/J mice to Klebsiella pneumoniae admi-nistered intratracheally: a pulmonary infection

model in an immunocompromised host. Com-

parative Medicine 53, 397–403Kent G, Iles R, Bear CE, et al. (1997) Lung disease in

mice with cystic fibrosis. Journal of Clinical

Investigation 100, 3060–9Lewko WM, Smith TL, Bowman DJ, Good RW,

Oldham RK (1995) Interleukin-15 and the growthof tumor derived activated T-cells. Cancer

Biotherapy 10, 13–20McCray Jr PB, Zabner J, Jia HP, Welsh MJ, Thorne PS

(1999) Efficient killing of inhaled bacteria inDeltaF508 mice: role of airway surface liquidcomposition. American Journal of Physiology 277,L183–90

McMorran BJ, Palmer JS, Lunn DP, et al. (2001)G551D CF mice display an abnormal host responseand have impaired clearance of Pseudomonas lungdisease. American Journal of Physiology. Lung

Cellular and Molecular Physiology 281, L740–7Morissette C, Francoeur C, Darmond-Zwaig C,

Gervais F (1996) Lung phagocyte bactericidalfunction in strains of mice resistant and suscep-tible to Pseudomonas aeruginosa. Infection and

Immunity 64, 4984–92Nelson A, Guilbault C, Radzioch D (2003) Animal

models used for the study of cystic fibrosis. In:Recent Research Developments in Cell Research.Research Signpost; Kerala, India, 1, 91–130

O’Neal WK, Hasty P, McCray Jr PB, et al. (1993) Asevere phenotype in mice with a duplication ofexon 3 in the cystic fibrosis locus. Human

Molecular Genetics 2, 1561–9Sadikot RT, Christman JW, Blackwell TS (2004)

Molecular targets for modulating lung inflamma-tion and injury. Current Drug Targets 5, 581–8

Schook LB, Carrick Jr L, Berk RS (1977) Experimentalpulmonary infection of mice by tracheal intubationof Pseudomonas aeruginosa: the use of antineo-plastic agents to overcome natural resistance.Canadian Journal of Microbiology 23, 823–6

Snouwaert JN, Brigman KK, Latour AM, et al. (1992)An animal model for cystic fibrosis made by genetargeting. Science 257, 1083–8

Sokol PA, Sajjan U, Visser MB, Gingues S, Forstner J,Kooi C (2003) The CepIR quorum-sensing systemcontributes to the virulence of Burkholderiacenocepacia respiratory infections. Microbiology149, 3649–58

Southern PM Jr, Mays BB, Pierce AK, Sanford JP (1970)Pulmonary clearance of Pseudomonas aeruginosa.Journal of Laboratory and Clinical Medicine 76,548–59

Starke JR, Edwards MS, Langston C, Baker CJ (1987) Amouse model of chronic pulmonary infection withPseudomonas aeruginosa and Pseudomonas cepa-

cia. Pediatric Research 22, 698–702Stotland PK, Radzioch D, Stevenson MM (2000)

Mouse models of chronic lung infection withPseudomonas aeruginosa: models for the study ofcystic fibrosis. Pediatric Pulmonology 30, 413–24



Urban TA, Griffith A, Torok AM, Smolkin ME, BurnsJL, Goldberg JB (2004) Contribution of Burkhol-deria cenocepacia flagella to infectivity andinflammation. Infection and Immunity 72,5126–34

van Doorninck JH, French PJ, Verbeek E, et al.(1995) A mouse model for the cystic fibrosisdelta F508 mutation. EMBO Journal 14,4403–11

van Heeckeren AM, Schluchter MD (2002) Murinemodels of chronic Pseudomonas aeruginosa lunginfection. Laboratory Animals 36, 291–312

Van Heeckeren AM, Tscheikuna J, Walenga RW, et al.(2000) Effect of Pseudomonas infection on weightloss, lung mechanics, and cytokines in mice.

American Journal of Respiratory and Critical Care

Medicine 161, 271–9Weksler B, Ng B, Lenert J, Burt M (1994) A simplified

method for endotracheal intubation in the rat.Journal of Applied Physiology 76, 1823–5

Yu H, Hanes M, Chrisp CE, Boucher JC, Deretic V(1998) Microbial pathogenesis in cystic fibrosis:pulmonary clearance of mucoid Pseudomonas

aeruginosa and inflammation in a mouse model ofrepeated respiratory challenge. Infection and Im-munity 66, 280–8

Zeiher BG, Eichwald E, Zabner J, et al. (1995) Amouse model for the delta F508 allele of cysticfibrosis. Journal of Clinical Investigation 96,2051–64



cystic fibrosis lung disease following infection with pseudomonas aeruginosa in cftr knockout mice...

Documents