Cystic fibrosis swine fail to secrete airway surface liquid in response to inhalation of pathogens

Abstract

Cystic fibrosis is caused by mutations in the gene encoding the cystic fibrosis transmembrane conductance regulator (CFTR) channel, which can result in chronic lung disease. The sequence of events leading to lung disease is not fully understood but recent data show that the critical pathogenic event is the loss of the ability to clear bacteria due to abnormal airway surface liquid secretion (ASL). However, whether the inhalation of bacteria triggers ASL secretion and whether this is abnormal in cystic fibrosis has never been tested. Here we show, using a novel synchrotron-based in vivo imaging technique, that wild-type pigs display both a basal and a Toll-like receptor-mediated ASL secretory response to the inhalation of cystic fibrosis relevant bacteria. Both mechanisms fail in CFTR^-/- swine, suggesting that cystic fibrosis airways do not respond to inhaled pathogens, thus favoring infection and inflammation that may eventually lead to tissue remodeling and respiratory disease.Cystic fibrosis is caused by mutations in the CFTR chloride channel, leading to reduced airway surface liquid secretion. Here the authors show that exposure to bacteria triggers secretion in wild-type but not in pig models of cystic fibrosis, suggesting an impaired response to pathogens contributes to infection.

Fig. 1

Synchrotron-based phase contrast imaging. a Schematic of the in vivo imaging set up. b Detection of the air/ASL layer interface in the lumen of swine trachea in vivo. When x-ray (undisturbed wavefront) pass through the trachea, the difference in refractive index between the ASL and the air in the lumen results in a phase shift of the x-ray (distorted wavefront), which is detected by phase contrast imaging (PCI). PCI cannot resolve the ASL/tissue interface because the ASL refractive index is very similar to that of the tissue. Thus we used agarose beads as “measuring rods” to determine the position of the tissue with respect to the air/ASL interface. In the phase contrast image obtained in vivo (grayscale image on right panel), the air/ASL interface is clearly observable as a dark line (highlighted with dotted line). The edge of the agarose bead is outlined by the dotted circle. The distance from the bottom of the agarose bead to air/ASL interface is used to measure the height of the ASL layer. Note that the air/ASL interface is below the level of the meniscus formed by the ASL on the agarose bead. Scale bar represents 500 µm. c Schematic diagram illustrating the forces acting on the agarose bead (surface tension and gravity) and the bead–tissue interaction. d Frequency distribution and mean (93 ± 8 µm, indicated by arrow) of ASL height in wild-type swine trachea exposed to bacteria-free agarose beads in vivo (n = 39 agarose beads obtained from 24 wild-type live pigs)

Fig. 2

CFTR^−/− swine fail to respond to Pseudomonas aeruginosa in vivo. Scatter plot and median (red line) of the effect of a bacteria-free and b bacteria-laden agarose beads on the ASL layer of wild-type swine. Bacteria-free beads trigger an ASL increase from 105 to 123 µm (P < 0.05, n = 10 from seven pigs, Wilcoxon match-pairs signed rank test) while bacteria-laden beads triggered a height increase from 117 to 168 µm (P < 0.05, n = 10 from six pigs, Wilcoxon match-pairs signed rank test). c Increase in ASL height triggered by bacteria-laden (n = 10) and bacteria-free (n = 10) agarose beads (P < 0.0001, t = 7.76, df = 18, Student’s t-test). Effect of d bacteria-free and e bacteria-laden agarose beads on the ASL layer of CFTR^−/− swine. Bacteria-free beads trigger an ASL increase from 172 to 178 µm (P < 0.05, n = 9 from six pigs, Wilcoxon match-pairs signed rank test) while bacteria-laden beads triggered a height increase from 150 to 158 µm (P < 0.05, n = 11 from eight pigs, Wilcoxon match-pairs signed rank test). f The increase in ASL height triggered by bacteria-laden (n = 11) and bacteria-free (n = 9) agarose beads were not significantly different in CFTR^−/− swine (P = 0.49, Mann–Whitney test). Data are presented as mean ± SEM and within each panel the columns labeled with an asterisk differ significantly

Fig. 3

Stimulation of pattern recognition receptors triggers ASL secretion in wild-type swine in vivo. a P. aeruginosa PAK strain-laden agarose beads (PAK) significantly stimulates ASL secretion. PAKΔfliC-laden agarose beads (ΔfliC), which lacks flagellin, induce much less ASL secretion. Flagellin-laden (10 ng/ml) agarose beads (FLA) triggered a similar response as PAK beads (n = 11 bacteria-free from six pigs, n = 10 PAK from six pigs, n = 10 ΔfliC from seven pigs, and n = 8 FLA from five pigs; P < 0.0001, F(3, 35) = 36.59, ANOVA and Tukey’s multiple comparison test). b S. aureus-(SA), H. influenza-laden (HI), and LPS (20 µg/ml)-laden agarose beads stimulate ASL secretion in wild-type swine (n = 10 bacteria-free from seven pigs, n = 7 S. aureus from four pigs, n = 7 H. influenzae from four pigs, and n = 7 LPS from five pigs; P = 0.0003, F(3, 27) = 8.79, ANOVA and Tukey’s multiple comparison test). c The non-pathogenic B. subtilis (BS)- and E. coli (EC)-laden agarose beads failed to trigger ASL secretion in wild-type swine (n = 10 bacteria-free from six pigs, n = 7 E. coli DH5-α from five pigs, and n = 7 B. subtilis from six pigs; P = 0.34, F(2, 21) = 0.40, ANOVA and Tukey’s multiple comparison test). Secretion assay data showed that d IL-6 (50 ng/ml, n = 16 preparations from seven CFTR^−/− animals and n = 9 preparations from four wild-type animals, P < 0.0001, F(3, 60) = 12.79, ANOVA and Tukey’s multiple comparison test) and e IL-8 (50 ng/ml, n = 15 for CFTR^−/− airway tissue from seven pigs and n = 10 for wild-type tissue from four pigs, P < 0.0001, F(3, 56) = 13.09, ANOVA and Tukey’s multiple comparison test) both trigger submucosal gland secretion in wild-type airway, but fail to stimulated ASL secretion in CF airway. Data is presented as mean ± SEM and within each panel the columns labeled with different letters differ significantly

Fig. 4

CF tissues fail to produce basal ASL secretion. a CFTR^−/− swine (CF, n = 9 from six animals) display lower basal ASL secretion than wild-type in vivo (n = 10 from seven animals, mean ± SEM, P = 0.001, Mann–Whitney test). b Wild-type ex vivo preparations with surgically removed submucosal glands (SMG) produced lower basal ASL (n = 12 with SMG from six preparations and n = 13 without SMG from six preparations, P < 0.0001, Mann–Whitney test). c Wild-type ex vivo preparations incubated with CFTRinh172 (100 µM) produced lower basal ASL than non-treated preparations (n = 6 from four preparations for control, and n = 14 from five preparations for CFTRinh172, P < 0.0001, Mann–Whitney test). d Topical application of lidocaine (Lido) had a small effect on basal ASL secretion. Simultaneous treatment with the CFTR inhibitor (Lido + inh172) completely blocked basal ASL secretion (n = 13 control from five preparations, n = 21 Lido from eight preparations, and n = 16 Lido + inh172 from seven preparations; P < 0.0001, F(2, 47) = 59.35, ANOVA and Tukey’s multiple comparison test). e Tetrodotoxin (1 µM, TTX) treatment had no effect on basal ASL secretion (n = 18 control from eight preparations, and n = 11 TTX from four preparations, P = 0.58, Mann–Whitney test). Atropine had no effect on basal ASL secretion in f ex vivo preparations (10 µM, n = 8 without atropine from three preparations and n = 8 with atropine from four preparations, P = 0.76, t = 0.30, df = 14, Student’s t test) or g in vivo wild-type swine (treated with 0.04 mg/kg IM 2–10 min following induction of anesthesia, n = 8 without Atropine from six animals, and n = 10 with Atropine from seven animals, P = 0.75, t = 0.32, df = 16, Student’s t test). h Treatment with SQ22536 (0.5 mM) blocked ASL secretion (n = 10 control from four preparations and n = 11 SQ22536 from four preparations; P = 0.0003, Mann–Whitney test). Data is presented as mean ± SEM and within each panel the columns labeled with an asterisk differ significantly

Fig. 5

Diagram of an agarose bead immersed in ASL. The diagram shows the force due to surface tension and the capillary forces acting on the bead, and the parameters relevant to Eq. (3). Where R is the radius of the agarose bead, γ is the surface tension of ASL, δ is the height of the ASL layer in the trachea, z ₀ is the meniscus depth of the ASL due to the capillary effect of liquid around the agarose bead, θ is the contact angle between the ASL and the agarose. Our analysis shows that under our experimental conditions, the net effect of gravity and surface tension push the beads against the epithelium without causing significant tissue displacement