Therapeutic Effects of Hyaluronic Acid in Bacterial Pneumonia in Ex Vivo Perfused Human Lungs

Abstract

Rationale: Recent studies have demonstrated that extracellular vesicles (EVs) released during acute lung injury (ALI) were inflammatory.Objectives: The current study was undertaken to test the role of EVs induced and released from severe Escherichia coli pneumonia (E. coli EVs) in the pathogenesis of ALI and to determine whether high-molecular-weight (HMW) hyaluronic acid (HA) administration would suppress lung injury from E. coli EVs or bacterial pneumonia.Methods:E. coli EVs were collected from the perfusate of an ex vivo perfused human lung injured with intrabronchial E. coli bacteria for 6 hours by ultracentrifugation and then given intrabronchially or intravenously to naive human lungs. One hour later, HMW HA was instilled into the perfusate (n = 5-6). In separate experiments, HMW HA was given after E. coli bacterial pneumonia (n = 6-10). In vitro experiments were conducted to evaluate binding of EVs to HMW HA and uptake of EVs by human monocytes.Measurements and Main Results: Administration of HMW HA ameliorated the impairment of alveolar fluid clearance, protein permeability, and acute inflammation from E. coli EVs or pneumonia and reduced total bacteria counts after E. coli pneumonia. HMW HA bound to E. coli EVs, inhibiting the uptake of EVs by human monocytes, an effect associated with reduced TNFα (tumor necrosis factor α) secretion. Surprisingly, HMW HA increased E. coli bacteria phagocytosis by monocytes.Conclusions: EVs induced and released during severe bacterial pneumonia were inflammatory and induced ALI, and HMW HA administration was effective in inhibiting the uptake of EVs by target cells and decreasing lung injury from E. coli EVs or bacterial pneumonia.

Keywords: acute lung injury; ex vivo perfused human lung; extracellular vesicles; hyaluronic acid.

Figure 1.

Schematic of the ex vivo perfused human lung. Human lungs rejected for clinical transplantation were perfused with Dulbecco’s modified Eagle medium containing 5% albumin with 100 ml fresh human whole blood at a rate of 250 ml/min. Once rewarmed to 37°C, the lung was ventilated using a Vt of 300 ml with 5 cm H₂O of positive end-expiratory pressure and respiratory rate 10 breaths/min in room air. For lungs with alveolar fluid clearance (AFC) > 10%, 10⁹ cfu of Escherichia coli K1 strain was instilled to the middle or lower lobe. At 6 hours, extracellular vesicles (EVs) were isolated from the perfusate (E. coli EVs). In separate naive human lungs with AFC > 10%/h, E. coli EVs collected from 400 ml of perfusate were given intrabronchially or intravenously to induce lung injury. In separate experiments with lungs with AFC < 10%, 10⁹ cfu E. coli was instilled into the middle or lower lung lobes to induce severe bacterial pneumonia. For all treatment groups, 1 mg of HMW HA was administered through the pulmonary artery 1 hour after injury. HA = hyaluronic acid; HMW = high-molecular-weight; LLL = left lower lung lobe; LUL = left upper lung lobe; PEEP = positive end-expiratory pressure; RML = right middle lung lobe; RR = respiratory rate; RUL = right upper lung lobe.

Figure 2.

Effect of high-molecular-weight (HMW) hyaluronic acid (HA) on lung injury induced by Escherichia coli extracellular vesicle (EV) instillation. Administration of HMW HA as therapy restored alveolar fluid clearance (AFC) rate and reduced inflammation in the injured alveolus. (A) Gross human lung image of extravasation of Evan’s blue into the injured alveolus after intrabronchial instillation of E. coli EVs. Representative histopathological images of lung tissue slice from control, intravenous (i.v.) E. coli EV, and i.v. E. coli EV plus i.v. HMW HA groups with hemasolin and eosin staining. Instillation of E. coli EVs intrabronchially or intravenously significantly reduced AFC rate, which was restored with HMW HA administration. (Data are mean ± SD, n = 5–6 per treatment group, and *P < 0.05 by ANOVA [Bonferroni].) (B) Although administration of HMW HA decreased median absolute neutrophils counts, there was no statistical difference. (Data are median with interquartile range, n = 5–6 per treatment group.) Instillation of E. coli EVs intrabronchially or intravenously significantly increased levels of TNFα (tumor necrosis factor α) and total HA, presumably low-molecular-weight HA, in the injured alveolus and perfusate, respectively. Administration of HMW HA reduced TNFα in the injured alveolus and further increased total HA in the perfusate. (Data are mean ± SD, n = 5–6 per treatment group, *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001 by ANOVA [Bonferroni].) (C) There were no consistent effects on pulmonary artery pressure, airway pressure, or perfusate Po₂ levels over 6 hours or between individual groups. (Data are median with interquartile range, n = 5–6 per treatment group.) P values and confidence intervals are shown in Table E1 in the online supplement. BALF = BAL fluid; PAP = pulmonary artery pressure.

Figure 3.

Characterization of extracellular vesicles (EVs) released during Escherichia coli pneumonia in the ex vivo perfused human lung. (A) Plasma EVs appeared as small membrane-bound vesicles by electron microscopy (scale bar, 200 nm). The arrows highlight the different sizes of EVs found. (B) NanoSight analyses revealed that >50% of the EVs were ∼145–205 nm in size with a mean EV size of 155 nm. (C) By flow cytometry, EVs isolated from the perfusate 6 hours after injury were predominantly from endothelial cells and platelets. In addition, 27 ± 6% expressed CD44, a marker of microvesicles, and 59 ± 20% expressed CD9, a marker of exosomes, at 6 hours. Data are expressed as mean ± SD, n = 3–4. The Blue B-H axes relate to detection with a 525/50-nm bandpass filter when excited with a blue laser (488 nm). The Blue A-H axis relates to detection with a 695/40-nm bandpass filter when excited with a blue laser (488 nm). FITC = fluorescein isothiocyanate; PerCP = peridinin-chlorophyll proteins; SSC = side scatter.

Figure 4.

Binding of Escherichia coli extracellular vesicles (EVs) to high-molecular-weight (HMW) hyaluronic acid (HA). (A and B) PKH26-labeled E. coli EVs significantly bound to HMW HA plated on glass slides compared with control (scale bars, 15 μm), which was inhibited with coincubation with anti-CD44 antibody (Ab) by fluorescence (A) and total TNFα (tumor necrosis factor α) levels in EVs (B). Total TNFα levels in EVs were used as a surrogate measure of EV attachment to the HA. (Data are mean ± SD, *P < 0.05 and **P < 0.01 by ANOVA [Bonferroni], n = 4–8.) P values and confidence intervals are shown in Table E1.

Figure 5.

Update of Escherichia coli extracellular vesicles (EVs) by human monocytes. (A) The normal uptake of PKH26-labeled E. coli EVs by human monocytes was inhibited by coincubation with high-molecular-weight (HMW) hyaluronic acid (HA) by fluorescent microscopy as quantified by the average fluorescence intensity of PKH26 for each cell (scale bars, 50 μm). (B) Compared with EVs released from healthy lungs, E. coli EVs contained significantly higher mRNA levels of the inflammatory cytokines TNFα (tumor necrosis factor α) and IL-6 at 6 hours. (C) Coincubation of human monocytes with HMW HA decreased the release of TNFα and IL-6 caused by E. coli EVs by 29% and 53%, respectively, by human monocytes. Data are mean ± SD for PKH26 E. coli EV uptake and TNFα and IL-6 levels. Data are median with interquartile range for PCR data. *P < 0.05, **P < 0.01, and ****P < 0.0001 by ANOVA (Bonferroni) in C, ANOVA with Sidak’s multiple comparison test in A, and Mann-Whitney test in B; n = 3–12. P values and confidence intervals are shown in Table E1. qPCR = quantitative PCR.

Figure 6.

Therapeutic effects of high-molecular-weight hyaluronic acid (HMW HA) in Escherichia coli pneumonia. Intrabronchial instillation of E. coli bacteria caused severe lung injury, which was ameliorated with administration of HMW HA as therapy. (A) HMW HA reduced cellularity, interstitial thickening, and edema by hemasolin and eosin staining, restored alveolar fluid clearance (AFC) rate, and reduced lung protein permeability and pulmonary edema. HMW HA also reduced total E. coli cfu counts in the injured alveolus and increased the phagocytosis of bacteria by human blood monocytes in vitro. Data are mean ± SD for AFC rate, change in lung weight, and permeability, and data are median with interquartile range for BALF cfu counts, n = 6–10 per treatment group. *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001 by ANOVA (Bonferroni) or Mann-Whitney test. (B) Coincubation of human monocytes with HMW HA decreased bacterial cfu counts in the supernatant and increased GFP-labeled E. coli bacteria phagocytosis as measured by fluorescence. Data are mean ± SD, n = 5–12. **P < 0.01 and ****P < 0.0001 by ANOVA (Bonferroni). P values and confidence intervals are shown in Table E1. BALF = BAL fluid.

Figure 7.

Effect of Escherichia coli pneumonia with or without high-molecular-weight hyaluronic acid (HMW HA) administration on inflammation and pulmonary arterial pressure, airway pressure, and Po₂ levels. (A) There was no significant effect of HMW HA on total white blood cell (WBC) and neutrophil counts. However, HMW HA administration reduced the median level of TNFα (tumor necrosis factor α) in the BAL fluid (BALF) by 23% induced by E. coli pneumonia, although it was not statistically significant. Data are median with interquartile range, n = 6–10 per treatment group. *P < 0.05, **P < 0.01, and ****P < 0.0001 by Kruskal-Wallis with Dunn’s correction. P values and confidence intervals are shown in Table E1. (B) Although at individual time points there were statistically significant differences between groups, there appeared to be no overall effect of HMW HA administration on pulmonary arterial pressure, airway pressure, or Po₂ levels over time. Data are median with interquartile range, n = 6–10 per treatment group, P is significant versus corresponding group at each time point by Mann-Whitney test. PAP = pulmonary artery pressure.

Figure 8.

Effect of Escherichia coli pneumonia with or without high-molecular-weight hyaluronic acid (HMW HA) administration on total HA concentration in the injured alveolus and perfusate. (A) Administration of HMW HA dramatically increased total HA levels in both the injured alveolus and perfusate over time. Data are mean ± SD, n = 3–6. For BALF HA levels, *P < 0.05, **P < 0.01, and ****P < 0.0001 by ANOVA (Bonferroni). For perfusate serial data, P is significant versus corresponding group at each specific time point by Student’s t test. P values and confidence intervals are shown in Table E1. (B) In the treatment group, the majority of HA in both the injured alveolus and perfusate were HMW (>1,000 kD) in size despite 6 hours of perfusion. Even in the injured group, endogenous levels of HMW HA were elevated, suggesting some attempt at repair and recovery, although lower when compared with the treatment group. Size distribution of HA was separated out by gel filtration chromatography (Figure E1), and the concentration measured by ELISA. Data are mean ± SD, n = 3. BALF = BAL fluid; BSA = bovine serum albumin; Conc. = concentration.