Acid Ceramidase Rescues Cystic Fibrosis Mice from Pulmonary Infections

Previous studies have shown that sphingosine kills a variety of pathogenic bacteria, including Pseudomonas aeruginosa and Staphylococcus aureus. Sphingosine concentrations are decreased in airway epithelial cells of cystic fibrosis (CF) mice, and this defect has been linked to the infection susceptibility of these mice. Here, we tested whether the genetic overexpression of acid ceramidase rescues cystic fibrosis mice from pulmonary infections with P. aeruginosa. ABSTRACT Previous studies have shown that sphingosine kills a variety of pathogenic bacteria, including Pseudomonas aeruginosa and Staphylococcus aureus. Sphingosine concentrations are decreased in airway epithelial cells of cystic fibrosis (CF) mice, and this defect has been linked to the infection susceptibility of these mice. Here, we tested whether the genetic overexpression of acid ceramidase rescues cystic fibrosis mice from pulmonary infections with P. aeruginosa. We demonstrate that the transgenic overexpression of acid ceramidase in CF mice corresponds to the overexpression of acid ceramidase in bronchial and tracheal epithelial cells and normalizes ceramide and sphingosine levels in bronchial and tracheal epithelial cells. In addition, the expression of β1-integrin, which is ectopically expressed on the luminal surface of airway epithelial cells in cystic fibrosis mice, an alteration that is very important for mediating pulmonary P. aeruginosa infections in cystic fibrosis, is normalized in cystic fibrosis airways upon the overexpression of acid ceramidase. Most importantly, the overexpression of acid ceramidase protects cystic fibrosis mice from pulmonary P. aeruginosa infections. Infection of CF mice or CF mice that inhaled sphingosine with P. aeruginosa or a P. aeruginosa mutant that is resistant to sphingosine indicates that sphingosine and not a metabolite kills P. aeruginosa upon pulmonary infection. These studies further support the use of acid ceramidase and its metabolite sphingosine as potential treatments of cystic fibrosis.

nosa (2). However, the concept of altered mucociliary clearance as a cause of increased infection susceptibility was difficult to prove in vivo (3).
We and others have recently reported that the lipid composition of CF airway epithelial cells is significantly altered, with an increase of ceramide and a decrease of sphingosine in the luminal membrane of tracheal and bronchial epithelial cells (4)(5)(6)(7)(8)(9)(10)(11)(12)(13)(14)(15)(16). We have previously shown that sphingosine is central for killing bacterial pathogens in cystic fibrosis airways and that the decrease of sphingosine in cystic fibrosis airway epithelial cells is central for the increased infection susceptibility of cystic fibrosis mice or human cells isolated from cystic fibrosis patients (11)(12)(13). We and others also demonstrated that sphingosine kills a variety of bacterial species in vitro and in vivo, including P. aeruginosa, Escherichia coli, Haemophilus influenzae, Moraxella catarrhalis, Burkholderia cepacia, Staphylococcus aureus, Acinetobacter baumannii, and Porphyromonas gingivalis (11,12,(17)(18)(19)(20). Luminal ceramide concentrations are very low in healthy epithelial cells of the airways and increased in cystic fibrosis epithelial cells. This accumulation of ceramide results in the ectopic expression and clustering of ␤1integrins that further support the infection susceptibility of cystic fibrosis cells (13).
Thus, the balance of ceramide and sphingosine plays a central role in bacterial infections of cystic fibrosis cells. The concentrations of the two lipids in airway epithelial cells seem to be mainly regulated by acid ceramidase, as indicated by studies applying acid ceramidase in the lungs of cystic fibrosis mice (11,13), although the significance of endogenous acid ceramidase for bacterial infections of cystic fibrosis mice remains to be determined.
To analyze the role of acid ceramidase in cystic fibrosis, we generated mice that lack functional Cftr and overexpress acid ceramidase (mouse gene, Asah1). We then determined alterations of lipids in these mice and, most importantly, the infection susceptibility of these mice in comparison to Cftr-deficient mice. Using a sphingosine-resistant P. aeruginosa mutant, we further analyzed whether sphingosine or a metabolite mediates the killing of P. aeruginosa.
We demonstrate that the overexpression of acid ceramidase in cystic fibrosis mice normalizes ceramide and sphingosine concentrations in cystic fibrosis epithelial cells. This results in a normalization of the ectopic expression of ␤1-integrins in epithelial cells. Most importantly, the overexpression of acid ceramidase was sufficient to prevent pulmonary infections of cystic fibrosis mice with P. aeruginosa. Wild-type mice rapidly cleared wild-type P. aeruginosa upon pulmonary infection but were unable to eliminate a sphingosine-resistant P. aeruginosa strain. Cystic fibrosis mice were highly sensitive to pulmonary infection with the wild-type strain or a sphingosine-resistant mutant of P. aeruginosa. Inhalation of sphingosine restored the resistance of CF mice against wildtype P. aeruginosa strains but had no effect on the sphingosine-resistant P. aeruginosa mutant.

RESULTS
To define the role of acid ceramidase in airway epithelial cells for the increased infection susceptibility of cystic fibrosis mice, we crossed Cftr-deficient (CF) mice with acid ceramidase transgenic mice in order to overexpress acid ceramidase in cystic fibrosis mice. This novel mouse strain was named CF-Asah1tg. The activity of acid ceramidase in the trachea of CF-Asah1tg mice was higher than that in wild-type mice but lower than the activity of the acid ceramidase in control Asah1 transgenic mice (Fig.  1A). Syngeneic CF mice showed a reduced activity of the acid ceramidase in the trachea (Fig. 1A). More importantly, the activity of the acid ceramidase was also increased in tracheal epithelial cells of CF-Asah1tg mice compared to CF mice. The activity of the enzyme in CF tracheal epithelial cells of CF mice was reduced compared to that in wild-type controls (Fig. 1B). Note that the activity of the enzyme was determined in situ in tracheal epithelial cells (Fig. 1B). The activity measurements corresponded to the protein expression levels of the acid ceramidase in tracheal and bronchial epithelial cells, which were reduced in CF cells and restored in CF-Asah1tg mice compared to wild-type mice ( Fig. 1C and D). Activity of acid ceramidase determined in tracheal extracts from syngeneic wild-type (WT), CF, Asah1tg, and CF-Asah1tg mice. Acid ceramidase activity is greatly reduced in CF mice and increased to wild-type levels in the trachea of CF mice that are transgenic for acid ceramidase (CF-Asah1tg). Shown are the means Ϯ SD from 6 independent experiments. *, P Ͻ 0.05; ***, P Ͻ 0.001 (by (Continued on next page) Acid Ceramidase and Cystic Fibrosis Infection and Immunity Acid ceramidase generates sphingosine from ceramide. Sphingosine levels are decreased in CF epithelial cells (11)(12)(13). We therefore determined the concentration of sphingosine in isolated tracheal epithelial cells from wild-type, CF, and CF-Asah1tg mice. The results confirm the reduction of sphingosine in CF mice compared to wild-type mice ( Fig. 2A). The transgenic overexpression of acid ceramidase in CF mice restored sphingosine levels in the trachea of these mice compared to CF mice ( Fig. 2A). Sphingosine levels in CF-Asah1tg mice did not differ from sphingosine levels in control wild-type mice. In situ kinase assays that allow quantitative measurements of surface sphingosine in tracheal epithelial cells also revealed an increase of surface sphingosine in CF-Asah1tg mice compared to CF mice (Fig. 2B). This indicates that the expression of the transgene is added to the expression of endogenous acid ceramidase, which is reduced in CF mice. The data from the biochemical studies were confirmed by confocal fluorescence microscopy studies that also showed a marked increase of sphingosine on the surface of epithelial cells of CF-Asah1tg mice compared to CF mice (Fig. 2C).
Previous studies demonstrated that the application of acid ceramidase to CF mice also normalizes ceramide levels in epithelial cells of the airways. Thus, to further characterize CF-Asah1tg mice, we measured total ceramide in the lungs and tracheas of wild-type, CF, and CF-Asah1tg mice. The results confirm the function of the transgene in CF-Asah1tg mice and demonstrate that the transgenic expression of Asah1 normalized ceramide levels compared to CF mice ( Fig. 3A and B). CF mice showed increased levels of ceramide in the lung and trachea ( Fig. 3A and B). Likewise, we determined surface ceramide by in situ kinase assays of the tracheal surface and detected that the transgenic expression of the acid ceramidase also normalized ceramide on the surface of tracheal epithelial cells (Fig. 3C). Ceramide levels were increased in CF mice compared to wild-type mice. The effect of the transgenic expression of acid ceramidase on ceramide levels in airway epithelial cells was also demonstrated in confocal microscopy studies that revealed the normalization of ceramide specifically in bronchial epithelial cells of CF-Asah1tg mice compared to wild-type and CF mice (Fig. 3D).
We have shown that increased ceramide levels result in the ectopic expression of ␤1-integrin on the surface of bronchial epithelial cells in CF airways, while ␤1-integrin is absent from the luminal surface of wild-type airways (13). The ectopic expression of ␤1-integrin is critically important for the increased susceptibility of CF mice to infections with P. aeruginosa or S. aureus as evidenced in experiments in which we blocked ␤1-integrin (13). This blockade prevented infections of CF mice with pathogens (13). Thus, we tested whether the transgenic expression of acid ceramidase in CF mice also normalizes the expression of ␤1-integrin. Confocal microscopy studies demonstrated the ectopic expression of ␤1-integrin in CF epithelial cell layers, which was abrogated in the epithelial cell layers of CF-Asah1tg mice (Fig. 4). The pattern of ␤1-integrin expression in these mice corresponded to that in wild-type mice.
Acute and chronic infections with P. aeruginosa and S. aureus are some of the most important problems of CF patients (1,21,22). Thus, we determined the infection susceptibility of CF-Asah1tg mice. To this end, we infected wild-type, CF, Asah1tg, and CF-Asah1tg mice intranasally with three different P. aeruginosa strains, a septic, a laboratory, and a mucosal one, and determined the number of bacteria in the lungs of  Infection and Immunity these animals and the score of sickness of the mice directly before lung resection. CF mice were very susceptible to the pathogens, and the lungs already contained high numbers of bacteria a few hours after intranasal application, while wild-type and Asah1tg mice were almost completely resistant to infection ( Fig. 5A to C). The transgenic overexpression of acid ceramidase in CF-Asah1tg mice restored the resistance of CF mice to P. aeruginosa. Infection resulted in very low numbers of the pathogens in the lungs of these mice, very similar to the bacterial numbers in wild-type mice ( Fig. 5A to C), and almost abrogated the clinical symptoms typical of pneumonia ( Fig. 5A to C). The product of the activity of acid ceramidase is sphingosine. Thus, we aimed to determine whether sphingosine or a metabolite mediates the killing of P. aeruginosa in the lung. To this end, we generated a P. aeruginosa clone from ATCC 27853 that is resistant to sphingosine. This was achieved by repeated incubation with 40 M sphingosine. We obtained 5 different clones and used one of these clones in the present studies. A dose-response curve with sphingosine shows that the strain exhibits highlevel resistance to sphingosine (Fig. 6A). We infected wild-type mice that had high concentrations of sphingosine in their airways (13) with the wild-type ATCC 27853 and sphingosine-resistant (ATCC 27853resSPH) strains and determined the development of pneumonia. Wild-type mice rapidly cleared the wild-type ATCC 27853 strain from the lung, while they developed severe pneumonia upon infection with the sphingosineresistant strain (Fig. 6B). Next, we infected CF mice and CF mice that were treated by sphingosine inhalation prior to infection with wild-type P. aeruginosa ATCC 27853 and the sphingosine-resistant P. aeruginosa strain. Cystic fibrosis mice should be sensitive to both strains because sphingosine concentrations in their airways are low. If sphingosine is important for defense against P. aeruginosa, inhalation of sphingosine in CF mice should rescue these mice from infection with wild-type P. aeruginosa strain ATCC 27853, while inhalation should have no or a reduced effect on infection with the sphingosine-resistant strain. If a metabolite of sphingosine mediates the killing of P. aeruginosa in the lung, inhalation of sphingosine should have comparable effects on both strains.
The results showed that infection of CF mice with P. aeruginosa wild-type strain ATCC 27853 or the sphingosine-resistant strain ATCC 27853resSPH resulted in severe pulmonary infection (Fig. 6C). Inhalation of sphingosine rescued CF mice from infection with P. aeruginosa wild-type strain ATCC 27853. In contrast, and most importantly, inhalation of sphingosine did not rescue CF mice from infection with the sphingosineresistant strain (Fig. 6C).
To test whether patients with cystic fibrosis are infected with P. aeruginosa strains that are sensitive or resistant to sphingosine, we tested the sensitivities of a panel of mucoid and nonmucoid P. aeruginosa strains isolated from CF patients. All of these strains were sensitive to sphingosine (Fig. 6D), although all of them showed resistance to many clinically used antibiotics (Table 1).

DISCUSSION
Here, we demonstrate that the transgenic overexpression of acid ceramidase in CF mice is sufficient not only to normalize ceramide and sphingosine levels but also to normalize functional consequences such as the ectopic expression of ␤1-integrin and, most importantly, to restore resistance to pulmonary P. aeruginosa infections. These studies further support a key role of acid ceramidase and sphingosine in the pathophysiology of CF. Studies employing a sphingosine-resistant P. aeruginosa strain support the notion that sphingosine is key to the pulmonary elimination of P. aeruginosa and that this defense mechanism is defective in CF mice. We have previously shown the reduced expression of acid ceramidase in tracheal and bronchial epithelial cells of CF mice but also in the lungs of CF patients and in liquid airway cultures of human epithelial cells (11)(12)(13). We also demonstrated that the application of recombinant acid ceramidase normalized ceramide and sphingosine levels and prevented infections with P. aeruginosa and S. aureus. However, these studies The results show that the transgenic expression of acid ceramidase in CF mice normalizes the increased ceramide levels in airway epithelial cells of CF mice and reduces the concentration of ceramide in tracheal and bronchial epithelial cells of CF mice to levels that are comparable to the ceramide levels in wild-type mice. Confocal microscopy studies were performed with Cy3-coupled anti-ceramide antibodies. Fluorescence was also quantified (D). Given are the means Ϯ SD or a representative result from 6 independent studies each. ***, P Ͻ 0.001 (by ANOVA and post hoc Student's t test). used a pharmacological approach applying acid ceramidase exogenously and do not prove that endogenous acid ceramidase is important for resistance to bacterial pneumonia. Thus, we aimed to provide genetic proof of the protective role of endogenous acid ceramidase in bacterial infections of the airways.
The transgenic overexpression of acid ceramidase results in the normalization of sphingosine levels in respiratory epithelial cells. Sphingosine kills bacterial pathogens on epithelial surfaces and the skin (11)(12)(13)(17)(18)(19)(20). We have recently shown that sphingosine binds to cardiolipin, clusters the lipid, and induces the formation of large and very rigid domains (20). The alterations of the biophysical properties of the plasma membrane result in leakiness of the plasma membrane, rapid depolarization, and killing of the bacteria (20). Since sphingosine must directly contact and even enter the bacteria, it seems to be important that the overexpression of acid ceramidase restores the surface levels of sphingosine, which comes into direct contact with invading pathogens. In addition, cells might release microparticles containing sphingosine, and the sphingosine concentration in these particles very likely also depends on the expression levels of acid ceramidase.
At present, it is unknown how acid ceramidase is downregulated in CF cells. To overcome this gap, we used the approach of overexpressing acid ceramidase. In this approach, acid ceramidase expression is controlled by the CAG promoter and does not underlie the endogenous regulation of the expression of acid ceramidase.
Previous studies have indicated that inhalation of recombinant acid ceramidase normalizes the expression of ceramide, sphingosine, and ␤1-integrin and, most importantly, also infection with P. aeruginosa. The present study demonstrates that normalization of the endogenous expression of acid ceramidase in CF airways also normalizes these pathologies and protects CF mice from severe pulmonary infection with P. aeruginosa. Clinically, bacterial lung infections constitute the major problem for most patients with CF (1,21,23). At present, gastrointestinal symptoms can be well controlled, but it is still very difficult and often impossible to prevent acute and finally chronic pneumonia. Chronic pneumonia, which is caused by P. aeruginosa in most CF patients, is also often the reason for the progressive destruction of the lung and shortened life expectancy in these patients. Moreover, chronic P. aeruginosa infections are also often observed in patients with chronic obstructive pulmonary disorder (COPD) (22). Ceramide has been shown to be increased in the lungs of COPD patients (24,25), and COPD might be another disease in which acid ceramidase might be downregulated. Therefore, collectively, these studies strongly support the notion that treatment of CF patients with acid ceramidase or sphingosine may be very beneficial and may protect patients from the development of pulmonary infections.

MATERIALS AND METHODS
Mice. B6.129P2(CF/3)-Cftr TgH(neoim)Hgu (CF MHH ) congenic mice were generated as previously described (4,26,27). Briefly, Cftr TgH(neoim)Hgu mutant mice, which were generated by insertional mutagenesis in exon 10 of the Cftr gene, were inbred to create the congenic Cftr MHH strain. These mice were then backcrossed for more than 20 generations to a C57BL/6 background. Because these mice still express low levels of Cftr, they can be fed a standard mouse diet. They exhibit normal development but also display pulmonary pathology typical of CF (4-6, 28). Syngeneic C57BL/6 mice were used as wild-type controls.
Cftr MHH mice were crossed with mice transgenic for Asah1 to overexpress the acid ceramidase in mice lacking Cftr. Asah1 transgenic mice were also on a C57BL/6 background. The overexpression of the acid ceramidase was achieved by introducing an expression cassette in which Asah1 is expressed under the control of the CAG promoter into the mouse genome. Mice were cloned, and the integration of the transgene was tested by PCR. The overexpression of acid ceramidase was confirmed by measuring acid ceramidase activity in several tissues.
Asah1 transgenic (Asah1tg0) mice served as controls. We used female and male mice at an age of at least 16 weeks and weighing between 25 and 35 g. Mice were divided into cages of equal sizes (usually 3 to 4 mice) by animal unit technical staff with no involvement in study design. Cages were randomly assigned to an experimental group. The investigators were blind to group allocation during the experiment and/or when assessing the outcome.
Mice were housed and bred within isolated and ventilated cages in the mouse facility of the University Hospital, University of Duisburg-Essen, Germany. They were repeatedly tested for a panel of common murine pathogens according to the 2002 recommendations of the Federation of European Laboratory Animal Science Associations (https://felasa.eu/working-groups/guidelines), including S. aureus. The mice were free of all pathogens. Animal procedures were approved by the Bezirksregierung Duesseldorf, Duesseldorf, Germany, and the IACUC committee Cincinnati.
Immunohistochemistry of mouse trachea and lungs. Mouse trachea and lungs were stained for acid ceramidase, sphingosine, or ceramide as previously described (11)(12)(13). Mice were sacrificed by The tissues were then secondarily labeled with Cy3-coupled anti-rabbit IgG (for anti-acid ceramidase antibodies), Cy3-coupled anti-mouse IgM (for anti-sphingosine or anti-ceramide antibodies), or anti-rat IgG (for anti-␤1-integrin antibodies) F(ab) 2 fragments (Jackson Immunoresearch) in H/S plus 1% FCS for 30 min. Sections were washed three times as described above and once with PBS and embedded in Mowiol. Samples were evaluated by confocal microscopy on a Leica TCS-SL confocal microscope equipped with a 40ϫ lens, and images were analyzed with Leica LCS software (Leica Microsystems, Mannheim, Germany). All comparative samples were measured at identical settings. Control stainings were performed with Cy3-labeled secondary antibodies only or with irrelevant rabbit IgG, mouse IgM, or rat IgG, followed by staining with the corresponding Cy3-labeled secondary antibodies. These controls revealed very weak staining and confirmed the specificity of the antibody staining. Fluorescence intensities were quantified using ImageJ.
Luminal surface levels of ceramide. The intact epithelial surface of the trachea was incubated with 0.01 U diacylglycerol (DAG) kinase (catalog number BML-SE100; Enzo) in a solution containing 4 l of 150 mM sodium acetate (pH 7.4), 1 mM ATP, and 10 Ci [␥-32 P]ATP for 15 min at 30°C. Controls were incubated with the same buffer without kinase or were left untreated. The kinase reaction was terminated by transferring the trachea into a solution containing CHCl 3 -CH 3 OH-1 N HCl (100:100:1, vol/vol/vol), 170 l of a buffered saline solution (135 mM NaCl, 1.5 mM CaCl 2 , 0.5 mM MgCl 2 , 5.6 mM glucose, 10 mM HEPES [pH 7.2]) and 30 l of 100 mM EDTA were added, and lipids were extracted, separated on silica G60 TLC plates, and analyzed by liquid scintillation counting and a standard curve of C 16 -to C 24 -ceramides, as previously described (11,13).

CF-5 (sputum, mucoid)
CF-6 (throat swab, nonmucoid) Piperacillin  I  I  I  I  R  I  I  Piperacillin-tazobactam  I  I  I  I  R  I Ciprofloxacin  R  I  I  I  I  I  R  Levofloxacin  R  I  I  I  I  I

CF-8 (sputum, mucoid)
The resistance of the clinical isolates to most clinically employed antibiotics was determined as described in the text. R, resistant; S, sensitive; I, susceptible after exposure to increased concentrations.
Luminal surface levels of sphingosine. Mice were sacrificed, and the trachea was removed, opened, and placed on a 30°C prewarmed plastic plate. We then added 4 l of kinase buffer consisting of 150 mM sodium acetate (pH 7.4), 0.001 U of sphingosine kinase 1 (catalog number 6068-SK-010; R&D, Minneapolis, MN, USA), 1 M ATP, and 10 Ci [␥-32 P]ATP onto the luminal surface of the trachea. Controls were incubated with the same buffer without sphingosine kinase. The kinase reaction was terminated by the addition of the trachea to 100 l H 2 O, followed by the addition of a solution containing 20 l 1 N HCl, 800 l CHCl 3 -CH 3  Ceramide measurements in lung extracts. Tracheae or lungs were shock frozen; lysed in a solution containing 1% Igepal, 25 mM Tris-HCl (pH 7.4), 125 mM NaCl, 10 mM EDTA, 10 mM sodium pyrophosphate, and 10 g/ml each aprotinin and leupeptin; and immediately sonicated with a tip sonicator to contain homogeneous lysates. Aliquots were extracted in CHCl 3 -CH 3 OH-1 N HCl (100:200:1, vol/vol/vol). The lower phase was dried and resuspended in a detergent solution (7.5% [wt/vol] n-octyl glucopyranoside and 5 mM cardiolipin in 1 mM DTPA). The kinase reaction was initiated by the addition of a solution containing 10 l DAG kinase (GE Healthcare Europe, Munich, Germany), 0.1 M imidazole-HCl (pH 6.6), 0.2 mM DTPA, 70 mM NaCl, 17 mM MgCl 2 , 1.4 mM EGTA, 2 mM dithiothreitol, 1 M ATP, and 5 Ci [␥-32 P]ATP. Samples were incubated for 60 min at 37°C with shaking (350 rpm) and processed as described above.
Infections. P. aeruginosa strain 762, ATCC 27858, a mucoid isolate from individuals with CF, P. aeruginosa strain 1242, or the sphingosine-resistant P. aeruginosa strain was grown overnight on tryptic soy agar (TSA) plates and resuspended in 40 ml prewarmed, sterile tryptic soy broth (TSB) (Becton, Dickinson Biosciences) in Erlenmeyer flasks at an optical density (OD) of 0.200 to 0.250. Bacteria were then grown in TSB for 60 min at 37°C with shaking at 125 rpm until they reached the early logarithmic phase. After this, bacteria were centrifuged (10 min at 1,710 ϫ g), the supernatant was removed, bacteria were washed with prewarmed RPMI 1640 plus 10 mM HEPES and resuspended in the same medium, and the concentration was determined by measurement of the OD and comparison to a calibration curve. Finally, the concentration of the bacteria was adjusted to 5 ϫ 10 6 CFU per 20 l in prewarmed RPMI 1640 medium plus 10 mM HEPES (4,13), and the inoculum was directly used (4,13). If indicated, mice inhaled 800 l of a 125 M sphingosine suspension in 0.9% NaCl 30 min prior to infection (the mice inhale less than 10% of the inhalation volume). To infect mice with the P. aeruginosa strains, mice were lightly anesthetized as previously described (4,29). The mice were then inoculated intranasally with 5 ϫ 10 6 CFU P. aeruginosa by employing a plastic-coated 30-gauge needle, which was inserted 2 mm into the nose. We determined the mouse health status 4 h after infection using the following scores: a sickness score of 0 for unaffected (healthy appearance), a score of 1 for slightly affected (ruffled fur), a score of 2 for moderately affected (ruffled fur, breathing slightly impaired, and normal body temperature), and a score of 3 for severely affected (ruffled fur, heavy breathing, and lower body temperature). Bacterial numbers were then determined in the mouse lungs 4 h after infection. To this end, mice were sacrificed, and the lungs were removed, homogenized, and lysed in 5 mg/ml saponin to release intracellular bacteria. Bacteria were then pelleted at 2,240 ϫ g, washed once in sterile PBS, diluted, plated, and grown in duplicate on TSA plates for 12 h. Bacterial numbers were counted. Infection experiments were approved by the Bezirksregierung Duesseldorf, Duesseldorf, Germany, under permission numbers 84-02.04.2013.A348 and 81-02.04.2019.A134 and by the IACUC committee Cincinnati. The actual care and treatment of the animals were performed and/or overseen by veterinarians of the Central Animal Facility of the University Hospital Essen, Essen, Germany.
Sphingosine treatment of P. aeruginosa in vitro. To test the antibacterial efficacy of sphingosine in killing different P. aeruginosa strains in vitro, we incubated 1 ϫ 10 3 bacteria in H/S with increasing amounts of sphingosine (C 18 -sphingosine; Avanti Polar Lipids, Inc., AL, USA). To this end, a sphingosine stock solution of 10 mM in 0.9% NaCl was prepared by sonication and stored at Ϫ20°C. The stock solution was bath sonicated directly before use, diluted in H/S, and sonicated again for 10 min. Bacteria were incubated with the indicated sphingosine concentrations for 1 h at 37°C on a horizontal shaker at 125 rpm. After this time, an aliquot of the bacteria was plated on Luria broth (LB) agar plates and incubated overnight to determine bacterial survival.
Generation of sphingosine-resistant P. aeruginosa strains. P. aeruginosa strain ATCC 27853 was prepared as described above, and 24 aliquots of 1,000 CFU each were incubated for 60 min with 40 M sphingosine in 500 l as described above. We then added 500 l TSB, and cultures were grown overnight. Resistant cultures were plated on TSB agar plates, and clones were picked after growth overnight. Clones were treated for 5 additional rounds with 40 M sphingosine and then used for experiments.
Clinical P. aeruginosa strains and antibiotic resistance measurements. All clinical samples were used after performing conventional microbiological diagnostics. The study did not include patients' details and did not result in additional constraints for the patients. All data were anonymously analyzed without the need for patient consent due to the retrospective nature of the study. All procedures and methods were carried out in accordance with approved guidelines.
In this study, eight clinical Pseudomonas aeruginosa strains were included ( Table 1). All isolates were derived from sputum samples or throat swabs from CF patients. Antibiotic susceptibility testing was performed with commercial methods (MicroScan WalkAway plus, panel NC83; Beckman Coulter, Brea, CA, USA). Susceptibility testing was performed according to the manufacturer's instructions. For interpretation, EUCAST breakpoints (version 10.0, 2020) were used.
Quantification and statistical analysis. Data are expressed as arithmetic means Ϯ standard deviations (SD). For the comparison of continuous variables from independent groups, we used Student's t test for two groups and one-way analysis of variance (ANOVA) for more than two groups followed by post hoc Student's t test for all pairwise comparisons applying Bonferroni correction for multiple testing.
The P values for the pairwise comparisons were calculated after Bonferroni correction. All values were normally distributed. The statistical details (numbers, means Ϯ SD, and tests) are given in the figure legends. The sample size planning for the continuous variables in in vivo infection experiments was based on two-sided Wilcoxon-Mann-Whitney tests (G*Power version 3.1.7 software of the University of Duesseldorf, Germany). Investigators were blind for histology experiments and animal identity as described above. All data were quantified using ImageJ and are expressed as arbitrary units. In each photo, 20 randomly chosen areas corresponding to the luminal membrane of 20 different cells were quantified and averaged with the values obtained in the other photos from the fluorescence microscopy studies. G*Power version 3.1.7 of the University of Duesseldorf, Germany, and ImageJ are public programs available online without charge.