N-glycosylation augmentation of the cystic fibrosis epithelium improves Pseudomonas aeruginosa clearance

Abstract

Chronic lung colonization with Pseudomonas aeruginosa is anticipated in cystic fibrosis (CF). Abnormal terminal glycosylation has been implicated as a candidate for this condition. We previously reported a down-regulation of mannose-6-phosphate isomerase (MPI) for core N-glycan production in the CFTR-defective human cell line (IB3). We found a 40% decrease in N-glycosylation of IB3 cells compared with CFTR-corrected human cell line (S9), along with a threefold-lower surface attachment of P. aeruginosa strain, PAO1. There was a twofold increase in intracellular bacteria in S9 cells compared with IB3 cells. After a 4-hour clearance period, intracellular bacteria in IB3 cells increased twofold. Comparatively, a twofold decrease in intracellular bacteria occurred in S9 cells. Gene augmentation in IB3 cells with hMPI or hCFTR reversed these IB3 deficiencies. Mannose-6-phosphate can be produced from external mannose independent of MPI, and correction in the IB3 clearance deficiencies was observed when cultured in mannose-rich medium. An in vivo model for P. aeruginosa colonization in the upper airways revealed an increased bacterial burden in the trachea and oropharynx of nontherapeutic CF mice compared with mice treated either with an intratracheal delivery adeno-associated viral vector 5 expressing murine MPI, or a hypermannose water diet. Finally, a modest lung inflammatory response was observed in CF mice, and was partially corrected by both treatments. Augmenting N-glycosylation to attenuate colonization of P. aeruginosa in CF airways reveals a new therapeutic avenue for a hallmark disease condition in CF.

Figure 1.

Percent difference of attached FITC-conjugated lectins, lens culinaris lectin (LcH) and concanavalin A (ConA) normalized to IB3 cells. IB3 treatments included pTR2-CB-Δ264 hCFTR, pTR2-CB-hMPI, pTR2-CB-SMN (control), or hypermannose media at 500-μM concentration and S9 cells treated with pTR2-U6-CFTRsiRNA or siRNA negative control. (A) Data collected from FITC-ConA and (B) data collected using FITC–LCA/LcH. Statistical analysis was done using a two-tailed, paired t test (*P < 0.05 analyzed against IB3; ^#P < 0.05 analyzed against S9). Data were collected from three separate trials.

Figure 2.

Difference in percentage of green fluorescent protein (GFP) fluorescence of host cells due to internalized GFP plus PAO1 strain shown as a percentage change compared with untreated IB3 cells. GFP fluorescence from ingested PAO1-GFP was detected by FACS analysis. (A) IB3 cells were treated with pTR2-CB-Δ264 hCFTR, pTR2-CB-hMPI, or pTR2-CB-SMN (control), and S9 cells were treated with pTR2-U6-CFTRsiRNA or siRNA negative control. (B) IB3 cells were treated with the indicated amount of mannose. Statistical analysis was done using a two-tailed, paired t test (*P < 0.05 analyzed against IB3; ^#P < 0.05 analyzed against S9). Data were collected from three separate trials.

Figure 3.

Change of intracellular PAO1-GFP from host cells after a 4-hour clearance period compared with initial internalized bacterial levels after infection, ingestion, and gentamicin treatment to kill extracellular PAO1-GFP, and the difference in host cell death from treated and untreated host cells after the same clearance period. (A and C) IB3 cells were treated with pTR2-CB-Δ264 hCFTR, pTR2-CB-hMPI, or pTR2-CB-SMN (control), and S9 cells were treated with pTR2-U6-CFTRsiRNA or siRNA negative control. (B and D) IB3 cells were cultured in indicated concentrations of mannose rich media. Statistical analysis was done using a two-tailed, paired t test (*P < 0.05 analyzed against IB3; ^#P < 0.05 analyzed against S9). Data were collected from three separate trials.

Figure 4.

Total mucoid Pseudomonas aeruginosa CFU collected from weekly oropharynx culture swabs from cftr mice that were infected for 2 weeks with the mucoid strain via drinking water and treated with (A) AAV5-CB-mMPI viral vector, with AAV5-CB-GFP as control (n = 6) or (B) hypermannose diet of 5 mg/ml in drinking water with hyperglucose diet as control (n = 4). Weekly percentage changes in weight compared with Week 0 weight of infected mice treated with (C) AAV5-CB-mMPI, with AAV5-CB-GFP as a control (n = 6), or (D) hypermannose diet of 5 mg/ml in drinking water with hyperglucose diet as a control (n = 4). P < 0.05 using one-way ANOVA for repeat sampling between groups.

Figure 5.

Cultured CFU from trachea homogenates of cftr mice infected with mucoid strain and treated with (A) AAV5-CB-mMPI viral vector, with AAV5-CB-GFP as a nontherapeutic control (n = 6) or (B) hypermannose diet of 5 mg/ml in drinking water, with hyperglucose diet as a non-therapeutic control (n = 4). Statistical analysis was done using a two-tailed, paired t test.

Figure 6.

Images of inflammatory conditions revealed by hematoxylin and eosin staining of lung sections from cftr mice infected for 2 weeks with the mucoid strain through the drinking water. Shown are cases, in nontherapeutic mice, of (A) bronchiectasis (star) and (B and C) multifocal, moderate inflammation (within circle [B]) and, in therapeutic mice, (D) healthy airways and (E and F) noninflammatory conditions.