Therapeutic effects of high molecular weight hyaluronic acid in severe Pseudomonas aeruginosa pneumonia in ex vivo perfused human lungs

Abstract

We previously reported that extracellular vesicles (EVs) released during Escherichia coli (E. coli) bacterial pneumonia were inflammatory, and administration of high molecular weight hyaluronic acid (HMW HA) suppressed several indices of acute lung injury (ALI) from E. coli pneumonia by binding to these inflammatory EVs. The current study was undertaken to study the therapeutic effects of HMW HA in ex vivo perfused human lungs injured with Pseudomonas aeruginosa (PA)103 bacterial pneumonia. For lungs with baseline alveolar fluid clearance (AFC) <10%/h, HMW HA 1 or 2 mg was injected intravenously after 1 h (n = 4–9), and EVs released during PA pneumonia were collected from the perfusate over 6 h. For lungs with baseline AFC > 10%/h, HMW HA 2 mg was injected intravenously after 1 h (n = 6). In vitro experiments were conducted to evaluate the effects of HA on inflammation and bacterial phagocytosis. For lungs with AFC < 10%/h, administration of HMW HA intravenously significantly restored AFC and numerically decreased protein permeability and alveolar inflammation from PA103 pneumonia but had no effect on bacterial counts at 6 h. However, HMW HA improved bacterial phagocytosis by human monocytes and neutrophils and suppressed the inflammatory properties of EVs released during pneumonia on monocytes. For lungs with AFC > 10%/h, administration of HMW HA intravenously improved AFC from PA103 pneumonia but had no significant effects on protein permeability, inflammation, or bacterial counts. In the presence of impaired alveolar epithelial transport capacity, administration of HMW HA improved the resolution of pulmonary edema from Pseudomonas PA103 bacterial pneumonia.

INTRODUCTION

Acute respiratory distress syndrome (ARDS), a devastating condition common in patients with respiratory failure in the intensive care units (ICU), is characterized by an acute inflammatory response to various etiologies such as bacterial and viral pneumonias, sepsis, aspiration, or trauma, leading to injury to the alveolar epithelium and endothelium (1). Although progress has been made in the treatment of ARDS, it is mainly based on supportive care with ventilatory interventions, including low tidal volume ventilation and high-positive end-expiratory pressure (PEEP), and adjuncts such as prone positioning, neuromuscular blockade, and extracorporeal membrane oxygenation (2). The mortality rate of moderate to severe ARDS still remains as high as 40% worldwide (1) and >58% in some developing countries (3). Therefore, innovative therapies are needed.

Bacterial pneumonia or pneumonia associated with sepsis is the most common cause of ARDS. In our previous publication, we found that extracellular vesicles (EVs) released into the plasma during Escherichia coli (E. coli) bacterial pneumonia in ex vivo perfused human lungs were inflammatory in nature and by themselves led to acute lung injury (ALI) when given intravenously or intratracheally into naïve lungs (4). Administration of high molecular weight hyaluronic acid (HMW HA) as therapy bound to these inflammatory EVs via receptors such as CD44 and suppressed lung inflammation and improved lung protein permeability, leading to reduced pulmonary edema and decreased alveolar bacterial load (4). Although promising, one major concern of the study was whether the therapeutic effects of HA could be used in other etiologies of bacterial pneumonias.

The current study was undertaken to determine whether HMW HA had a similar therapeutic effect on a common cause of bacterial pneumonia; Pseudomonas aeruginosa (PA) is one of the most frequent causative agents of severe nosocomial pneumonias in the ICU (5, 6), which can cause ARDS and secondary sepsis (7, 8). Using the ex vivo perfused human lung, we developed a severe pneumonia model using the PA103 bacterial strain and tested the therapeutic effects of HMW HA in both healthy and marginal human lungs.

METHODS

Common research techniques and additional details can be found in the Supplemental Materials and Methods (see https://doi.org/10.6084/m9.figshare.15161541).

Ex Vivo Perfused Human Lung Preparation

The preparation of the ex vivo perfused human lung has been extensively characterized by our research group (Supplemental Fig. S1; see https://doi.org/10.6084/m9.figshare.13473450) (4, 9). Alveolar fluid clearance (AFC), a measure of the ability of the lung to reabsorb pulmonary edema fluid, was first measured in the control upper lobe by the change in protein concentration of a bronchoalveolar lavage fluid (BALF) over 1 h. For marginal human lungs with 0 < AFC < 10%/h, 10⁸ colony forming units (CFU) of PA103, a strain of Pseudomonas aeruginosa, was instilled into the lower lobe. As therapy, 1 or 2 mg of HMW HA (Sodium Hyaluronate with MW 850 to 1,000 KDa, LifeCore Inc., Chaska, MN) was administered intravenously (IV) through the pulmonary artery 1 h after injury. For human lungs with an intact alveolar epithelium and endothelium as suggested by an AFC > 10%/h, 2 mg HMW HA IV was given 1 h after initiation of PA103 pneumonia. The AFC rate, the primary end point, was measured by the change in protein concentration of a normal saline instillate containing 5% BSA over 1 h. AFC was calculated using the equation: AFC(%/h) = (1 − Ci/Cf) × 100 (Ci = protein concentration at time 0 and Cf = protein concentration after 1 h) (4, 10). Thirty milliliter of perfusate was collected every hour until T6 h to isolate perfusate EVs released during pneumonia by ultracentrifugation; EVs were resuspended in PBS (10 μL of PBS per EVs from 1 mL of perfusate).

Isolation and Characterization of Plasma EVs

Plasma EVs were isolated using ultracentrifugation as previously described. Total perfusate following 6 h of injury was centrifuged at 2000 g for 10 min, followed by 10,000 g for 30 min to remove cellular debris or protein complexes. The final supernatant was then ultracentrifuged (Beckman Coulter Optima L-100XP 19 ultracentrifuge) at 100,000 g at 4°C for 1 h twice. After ultracentrifugation, PA103 EVs were resuspended with PBS (10 µL per 1 mL of perfusate) and stored at −80°C. PA103 EVs were characterized by four methods as recommended by the International Society of Extracellular Vesicles (11, 12): 1) transmission electron microscopy to evaluate size and ultrastructure; 2) flow cytometry to quantify and determine the size distribution of EVs using a fluorescent FluoSpheres Size Kit (Invitrogen) and to measure known surface markers; 3) nanoparticle tracking analyses (Malvern Inc., NS300 NTA), an optical particle-tracking device, to quantitate the size and number of vesicles; and 4) levels of intravesicular mRNA using qPCR.

Fluorescent Labeling and Analysis of Plasma EVs with Flow Cytometry

Perfusate EVs were labeled with PKH26 (Red Fluorescent Cell Linker Mini Kit, Sigma-Aldrich, St. Louis, MO). For flow cytometry, EVs derived from 1 mL of perfusate were resuspended with staining buffer (BD Biosciences, San Jose, CA) containing CD9-FITC (eBioscience Inc., San Diego, CA), CD44-FITC (BD Biosciences), CD3-Alexa Fluor 488 (Thermo Fisher Scientific), CD14-FITC (BD Biosciences), CD41-Alexa Fluor 488 (BioLegend, San Diego, CA), CD326-Alexa Fluor 488 (BioLegend), CD66b-PerCP-Cy5.5 (BD Biosciences) or CD31-Alexa Fluor 488 (BioLegend), and then incubated for 30 min at room temperature.

The fluorescent expression of stained perfusate EVs was detected by a BD FACSAria Fusion Special Order (SORP) cell sorter (BD Biosciences) with 100-nm nozzle and neutral density (ND) filter 1. The threshold was set on the SSC 200. For fluorescence detection, we used a 586/15 band pass filter for PKH26; 525/50 band pass filter for CD9-FITC, CD44-FITC, CD3-Alexa Fluor 488, CD14-FITC, CD41-Alexa Fluor 488, CD326-Alexa Fluor 488 and CD31-Alexa Fluor 488; and 695/40 band pass filter for CD66b-PerCP-Cy5.5. Collected data were analyzed by Diva software (BD Biosciences).

Measurement of Inflammatory Effects of Plasma EVs on Human Monocytes

Human blood monocytes were collected from whole blood from healthy donors (4). Human monocytes, 2.5 × 10⁵ cells/well, were then stimulated by 40 μL of EVs collected from the plasma from T0 to T6 h separately. After 6 h of incubation at 37°C, the supernatant was collected. For HMW or low molecular weight (LMW) HA (MW 40 KDa, LifeCore Inc.) treatment group, 40 μL EVs from T6 h were added in each well followed by 100 μg/mL HMW or LMW HA. TNFα levels in the supernatant were measured by an ELISA kit (Invitrogen).

Measurement of Phagocytosis of PA103 Bacteria by Human Monocytes and Neutrophils

Human monocytes and neutrophils were stimulated with 100 ng/mL LPS (Escherichia coli O111:B4, Sigma Aldrich) with or without 25 µg/mL HMW or LMW HA for 24 h. Then 10⁷ CFU of PA103 bacteria, previously incubated with human plasma, was added. Human neutrophils were collected from whole blood of healthy donors using MACSxpress separator and neutrophil isolation kit (Miltenyi Biotec Inc.) according to the manufacturer’s protocol. To assess the bacterial clearance in the supernatant, after 90 min of incubation at 37°C, the supernatant was collected, seeded on 5% sheep blood agar plates and kept at 37°C overnight (Teknova Inc., Hollister, CA). CFU levels were measured the next day. For assessing the intracellular bacterial count, after 60 min of incubation at 37°C, 100 µg/mL gentamicin was added for another 30 min. The cells were then dissolved with 1% Triton X-100, and the lysates collected and seeded on 5% sheep blood agar plates at 37°C overnight. CFU levels were also measured the next day.

Statistics

Data are expressed as means ± standard deviation (SD) if the data were normally distributed and median with interquartile range (IQR) if not. Comparisons between groups were made using Mann–Whitney or Dunn’s test following Kruskal–Wallis analysis for nonparametric data. All statistical analyses were performed using GraphPad Prism software (La Jolla, CA).

RESULTS

Baseline Demographic Data for Donor Research Human Lungs

The demographic and clinical data for human lungs used for experiments with PA103 bacteria and HMW HA are listed in Tables 1 and 2.

Table 1.

Donor demographic and clinical data for experiments with PA103 bacterial pneumonia and HMW HA (AFC < 10%)

	PA103 (n = 9)	PA103 + HMW HA IV 1 mg (n = 4)	PA103 + HMW HA IV 2 mg (n = 6)
Age, yr	38.3 ± 15.6	52.0 ± 6.1	49.7 ± 18.3
Male sex, %	7 (77.8)	4 (100)	5 (83.3)
$\text{P}{\text{a}}_{\text{O}}{}_{{}_2}$/$\text{F}{\text{I}}_{{\text{O}}_2}$ (mmHg, IQR 25%–75%)	260 [178, 330]	147 [111, 205]	208 [160, 266]
Cdyn, mL/cmH2O	31.5 ± 11.3	35.6 ± 15.8	35.8 ± 7.5
Murray LIS	1.6 ± 0.7	1.7 ± 0.9	1.6 ± 0.4
Ischemic time, h:min	33:04 ± 6:08	28:52 ± 1:24	31:50 ± 5:20

Values are means ± SD or median with interquartile range (IQR). AFC, alveolar fluid clearance; $\text{F}{\text{I}}_{{\text{O}}_2}$, fraction of inspired oxygen; HMW HA, high molecular weight hyaluronic acid; IV, intravenous; PA, Pseudomonas aeruginosa; $\text{P}{\text{a}}_{\text{O}}{}_{{}_2}$, arterial partial pressure of oxygen. There were no significant differences among groups in terms of age, male sex, $\text{P}{\text{a}}_{\text{O}}{}_{{}_2}$/$\text{F}{\text{I}}_{{\text{O}}_2}$ ratio, lung compliance, or lung injury score (LIS) before organ harvest as well as total cold ischemic time before the start of the experiment. Comparison between groups were made using ANOVA (Bonferroni correction) or Dunn’s test following Kruskal–Wallis analysis.

Table 2.

Donor demographic and clinical data for experiments with PA103 bacterial pneumonia and HMW HA (AFC > 10%)

	PA103 (n = 6)	PA103 + HMW HA IV 2 mg (n = 6)
Age, yr	49.0 ± 14.1	54.8 ± 17.5
Male sex, %	5 (83.3)	4 (66.7)
$\text{P}{\text{a}}_{\text{O}}{}_{{}_2}$/$\text{F}{\text{I}}_{{\text{O}}_2}$ (mmHg, IQR 25%–75%)	296 [251, 350]	272 [198, 407]
Cdyn, mL/cmH2O	41.2 ± 6.9	44.8 ± 8.3
Murray LIS	1.2 ± 0.5	1.1 ± 0.5
Ischemic time, h:min	36:23 ± 5:53	34:49 ± 9:36

Values are as means ± SD or median with interquartile range (IQR). AFC, alveolar fluid clearance; $\text{F}{\text{I}}_{{\text{O}}_2}$, fraction of inspired oxygen; HMW HA, high molecular weight hyaluronic acid; IV, intravenous; PA, Pseudomonas aeruginosa; $\text{P}{\text{a}}_{\text{O}}{}_{{}_2}$, arterial partial pressure of oxygen. There were no significant differences among groups in terms of age, male sex, $\text{P}{\text{a}}_{\text{O}}{}_{{}_2}$/$\text{F}{\text{I}}_{{\text{O}}_2}$ ratio, lung compliance or lung injury score (LIS) before organ harvest, as well as total cold ischemic time before the start of the experiment. Comparison between groups were made using Student’s t test or Mann–Whitney test.

HMW HA Treatment Improved AFC Rate and Decreased Lung Weight Gain and Protein Permeability in Human Lungs with an Initial AFC < 10%/h Injured with Severe PA103 Bacterial Pneumonia

Representative histological images from lung lobes injured with PA103 bacteria showed an increase in lung edema and cellularity compared with the control upper lung lobe at 6 h, which was partially improved with HMW HA administration. AFC decreased from median with IQR 4.8%/h (4.3, 6.8) in the control lung lobe to 0.0%/h (0.0, 2.2) after PA103 bacteria instillation. Subsequent IV administration of HMW HA 2 mg significantly restored and increased AFC rate to 12.8%/h (10.5, 14.7) compared with both the control and the PA103 injured groups and decreased total lung weight gain by 34% from median with IQR 824 g (666, 902) to 543 g (468, 642). Protein permeability significantly increased with PA103 bacteria pneumonia and was restored with HMW HA administration (Fig. 1A). Administration of HMW HA to perfused human lungs without bacterial injury numerically increased AFC rate whether baseline control AFC rate in the upper lobe was < or >10%/h (Supplemental Fig. S2; https://doi.org/10.6084/m9.figshare.15161508). There were no significant differences in total white blood cells (WBC) and absolute neutrophil counts in the BALF among all groups. Although not statistically significant, TNFα levels in the BALF decreased with treatment with either 1 or 2 mg HMW HA by 68% and 58%, respectively, compared with the PA103 group (Fig. 1B). There were no significant changes in IL-1β or IL-6 levels in the BALF (Supplemental Fig. S3A; see https://doi.org/10.6084/m9.figshare.15161517).

Figure 1.

Therapeutic effects of HMW HA on lung injury induced by PA103 instillation in human lungs with baseline AFC < 10%/h. Administration of HMW HA as therapy restored AFC rate and reduced inflammation in the injured alveolus. A: representative histopathological images of lung tissue slices from control, PA103 IT, and PA103 IT + HMW HA IV groups with H&E staining. Instillation of HMW HA IV significantly restored and increased AFC rate in a dose-dependent manner and numerically decreased total lung weight gain and protein permeability in human lungs injured by PA103 bacterial pneumonia. Data were presented as median with IQR, n = 3–9 per treatment groups. Comparisons between groups were made using Dunn’s test following Kruskal–Wallis analysis. B: total white blood cells (WBC), absolute neutrophil counts (ANC), and TNFα levels in BALF in human lungs injured by PA103 bacterial pneumonia with or without HMW HA therapy. There were no apparent effect of HMW HA on WBC and ANC in the BALF. However, HMW HA treatment numerically decreased TNFα levels in the BALF at 6 h. Data were presented as median with IQR, n = 4–9 per treatment groups. Comparisons between groups were made using Dunn’s test following Kruskal–Wallis analysis. C: messenger RNA levels for TNFα, IL-1β, and IL-6 in plasma EVs were measured using qPCR. Administration of HMW HA numerically decreased TNFα and IL-6 mRNA levels in plasma EVs at T1–3 h and T4–6 h vs. the corresponding control group, respectively. Data were presented as median with IQR, n = 3 per group. AFC, alveolar fluid clearance; BALF, bronchoalveolar lavage fluid; EV, extracellular vesicle; H&E, hematoxylin-eosin; HMW HA, high molecular weight hyaluronic acid; IQR, interquartile range; IV, intravenous; PA, Pseudomonas aeruginosa.

Effects of HMW HA Treatment on Proinflammatory mRNA Expression in Plasma EVs

Compared with T0, plasma EV TNFα mRNA levels increased over T1–3 h, IL-1β mRNA levels over T1–3 h, and IL-6 mRNA levels over T3–6 h in human lungs injured with PA103 pneumonia. Although not statistically significant, administration of HMW HA decreased plasma EV TNFα mRNA levels at T1–2 h by 54% and 53%, respectively, and IL-6 mRNA levels at T5–6 h by 78% and 85%, respectively, compared with the corresponding time groups. Plasma EV IL-1β mRNA levels increased significantly starting at T2 h compared with T0 h. However, there was no apparent effect of HMW HA on IL-1β mRNA levels (Fig. 1C).

Effects of HMW HA Administration on Bacterial Clearance and Respiratory and Hemodynamic Parameters in Human Lungs with Baseline AFC < 10%/h Injured with PA103 Bacterial Pneumonia

PA103 pneumonia significantly increased total bacterial CFU counts in the BALF [8.4 × 10⁶ CFU/mL (2.0 × 10⁶, 4.3 × 10⁷ CFU/mL) in PA103 pneumonia compared with 3.2 × 10⁵ CFU/mL, (3.0 × 10⁴, 1.3 × 10⁶ CFU/mL) in control group]; the uninjured control group had low levels of bacterial CFU growth even before the experiments. However, there were no significant differences in CFU counts after HMW HA administration compared with the PA103 group. In addition, HMW HA administration had no significant effects on CFU counts in the perfusate (Fig. 2A).

Figure 2.

Effect of HMW HA treatment on bacterial CFU counts and hemodynamic and respiratory changes in human lungs injured with PA103 bacteria with baseline AFC < 10%/h. A: HMW HA administration had no significant effects on total bacterial CFU levels in both the BALF or perfusate at 6 h. BALF of the uninjured control lung lobe had a low-baseline bacterial CFU counts, which may be the cause of the initial injury to the alveolar epithelium and endothelium before the experiments. Data were presented as median with IQR, n = 4–9 per treatment groups. Comparison between groups were made by using Dunn’s test following Kruskal–Wallis analysis. B: although treatment with HMW HA numerically increased lung compliance up to 220% at T6 h, there were no significant effects in pulmonary artery pressure, lung compliance, or perfusate Po₂ levels over 6 h or between individual groups. Data were presented as median with IQR, n = 4–9 per treatment groups. AFC, alveolar fluid clearance; BALF, bronchoalveolar lavage fluid; CFU, colony forming units; HMW HA, high molecular weight hyaluronic acid; IQR, interquartile range; PA, Pseudomonas aeruginosa.

During 6 h of perfusion, there were no significant differences in pulmonary artery pressure, lung compliance, or Po₂ among different timepoints or among individual groups (Fig. 2B). However, treatment with HMW HA numerically increased lung compliance over 6 h. At T6 h, treatment with 1 or 2 mg HMW HA increased lung compliance by 167% or 220%, respectively, compared with the corresponding PA103 pneumonia group.

Characterization of Plasma EVs in Human Lungs with Baseline AFC < 10%/h Injured with PA103 Bacterial Pneumonia

EVs released into the plasma during PA103 bacterial pneumonia appeared as spheroids; >50% of the EVs were <250 nm in size. By flow cytometry, median with IQR 32.7% (32.1, 32.9) of EVs at T6 h as a percentage of total PKH26-labeled EVs were CD9 positive, label for exosomes, which was increased by 152% compared with that at T0 h, and 10.9% (7.8, 15.0) of EVs as a percentage of total PKH26-labeled EVs were CD44 positive, label for microvesicles. At 6 h, most EVs were released from platelets (CD41+) and endothelial cells (CD31+). Others cellular sources included monocytes (CD14+) and lymphocytes (CD3+). We were unable to detect EVs released from neutrophils using the CD66b+ Ab or epithelial cells using the CD326+ Ab (Fig. 3A). The proportion of the size of EVs did not change dramatically over 6 h (Fig. 3B). Representative flow cytometry figures are shown for CD9 and CD44 to demonstrate the specificity of the fluorescent-labeled antibodies (Supplemental Fig. S4; see https://doi.org/10.6084/m9.figshare.15161526).

Figure 3.

Characterization of plasma EVs released during PA103 bacterial pneumonia in ex vivo perfused human lungs. A: plasma EVs appeared as small membrane bound vesicles by electron microscopy (bar = 200 nm). By flow cytometry, 32.7% of EVs expressed CD9, a marker of exosomes at T6 h, which was higher than T0 h, and 10.9% of EVs expressed CD44, a marker of microvesicles at T6 h. By flow cytometry, EVs isolated from the perfusate following 6 h of injury were predominantly from platelets and endothelial cells. Data were presented as median with IQR, n = 3. B: by NanoSight analyses, >50% of the EVs were approximately 50–250 nm in size at all timepoints. EVs, extracellular vesicles; IQR, interquartile range; PA, Pseudomonas aeruginosa.

Effects of HMW HA Treatment on Inflammation and Bacteria Phagocytosis by Human Monocytes and Neutrophils

To investigate the effects of plasma EVs from perfused human lungs injured with PA103 bacteria on inflammation, EVs from T0 to T6 h were added to human blood monocytes separately in vitro, and TNFα levels in the supernatant were measured after 24 h incubation. TNFα levels in the medium gradually increased in monocytes injured with plasma EVs, particularly at later timepoints. The increase in TNFα may reflect an increase in the inflammatory properties of the EVs, as suggested by Fig. 1C or an increase in the total number of vesicles over 6 h, as suggested by Fig. 3A or both. However, administration of EVs from T0 to T6 h from perfused lungs injured with PA103 bacteria and treated with HMW HA decreased the secretion of TNFα by the monocytes over 6 h. To corroborate the results, we found that in monocytes injured with EVs from perfused human lung injured with bacteria at T6 h and treated with exogenous HMW HA 100 µg/mL, the TNFα levels decreased significantly (Fig. 4A). Surprisingly, when treated with LMW HA 100 µg/mL, TNFα levels also decreased compared with that of monocytes injured with EVs at T6 h (Supplemental Fig. S5; see https://doi.org/10.6084/m9.figshare.15161529).

Figure 4.

Therapeutic effects of HMW HA on inflammation and PA103 bacterial phagocytosis by human monocytes and neutrophils in vitro. A: plasma EVs from perfused human lungs injured with PA103 bacterial pneumonia with or without HMW HA treatment from T0 to T6 h were added to human blood monocytes separately in vitro, and TNFα levels in the supernatant were measured following coincubation. Compared with monocytes injured with plasma EVs from perfused human lungs injured with bacterial pneumonia, secretion of TNFα by the monocytes was suppressed by treatment of plasma EVs from perfused human lungs injured with bacterial pneumonia and treated with HMW HA. In separate experiments, treatment of exogenous HMW HA at 100 μg/mL significantly decreased TNFα secretion in monocytes injured with plasma EVs (T6 h) from perfused human lungs injured with bacterial pneumonia. Data were presented as median with IQR, n = 4–6 per treatment groups, Comparison between groups were made with Mann–Whitney test. B: effects of HMW HA on phagocytosis of PA103 bacteria by human monocytes and neutrophils. HMW HA administration significantly increased bacteria phagocytosis by monocytes. Data were presented as median with IQR, n = 6–15 per treatment groups. Comparison between groups were made with Mann–Whitney. HMW HA administration also significantly increased bacteria phagocytosis and decreased inflammation by neutrophils. Data were presented as median with IQR, n = 12–14 per group. Comparison between groups were made with Mann–Whitney test. CFU, colony forming units; EVs, extracellular vesicles; HMW HA, high molecular weight hyaluronic acid; IQR, interquartile range; PA, Pseudomonas aeruginosa.

To determine the effect of HA on bacteria clearance, HMW or LMW HA was added to monocytes injured with LPS and PA103 bacteria. HMW HA administration significantly decreased PA103 CFU counts in the culture medium and increased the intracellular bacteria counts (Fig. 4B). There were no significant differences in PA103 CFU counts after LMW HA treatment (Supplemental Fig. S5), indicating that HMW HA rather than LMW HA enhanced the phagocytosis of PA103 bacteria. In addition, incubation of LPS injured neutrophils with HMW HA decreased bacterial levels in the culture medium and was associated with decreased secretion of TNFα by the neutrophils (Fig. 4B).

HMW HA Administration Improved AFC Rate Following Injury by PA103 Bacterial Pneumonia in Human Lungs with Baseline AFC > 10%/h

Intravenous HMW HA significantly increased AFC rate compared with PA103 injured lung lobes [median with IQR 5.3%/h (3.2, 7.1) after PA103 injury compared with 16.4%/h (13.5, 19.2) with HMW HA 2 mg administration]. Although there were no significant differences on lung weight gain, HMW HA numerically decreased protein permeability by 46% compared with PA103 pneumonia (Fig. 5A). There were no significant changes in absolute neutrophil counts or in TNFα levels in the BALF after HMW HA treatment. Although not statistically significant, both IL-1β and IL-6 levels in the BALF decreased with treatment with 2 mg HMW HA compared with the PA103 group (Supplemental Fig. S3B).

Figure 5.

Therapeutic effects of HMW HA on lung injury induced by PA103 bacterial instillation in human lungs with baseline AFC > 10%/h. A: instillation of HMW HA 2 mg significantly increased AFC rate in human lungs injured by PA103 bacterial pneumonia. Although there were no significant differences in lung weight gain, HMW HA treatment numerically improved protein permeability. Data were presented as medium with IQR, n = 4–6 per treatment groups. Comparison between groups were made using Dunn’s test following Kruskal–Wallis analysis. B: there were no significant differences in absolute neutrophil counts (ANC) and TNFα levels in BALF in human lungs injured by PA103 bacteria pneumonia treated with HMW HA at 6 h. Data were presented as median with IQR, n = 6 per treatment groups. C: HMW HA treatment had no significant effects in PA103 CFU counts in both the BALF and perfusate at 6 h in human lungs injured by PA103 bacteria. Data were presented as median with IQR, n = 6 per treatment groups. Comparison between groups were made using Dunn’s test following Kruskal–Wallis analysis. D: there were no significant effects in pulmonary artery pressure, lung compliance, or perfusate Po₂ levels over 6 h or between individual groups. Data were presented as median with IQR, n = 6 per treatment groups. Comparison vs. corresponding group by Mann–Whitney test. AFC, alveolar fluid clearance; BALF, bronchoalveolar lavage fluid; CFU, colony forming units; HMW HA, high molecular weight hyaluronic acid; IQR, interquartile range; PA, Pseudomonas aeruginosa.

Effects of HMW HA Administration on PA103 Bacterial Clearance and Respiratory and Hemodynamic Parameters in Human Lungs with Baseline AFC > 10%/h

Intravenous administration of HMW HA did not significantly change total bacterial CFU counts in the BALF or the perfusate (Fig. 5C). There were no significant effects of HMW HA administration on pulmonary artery pressure, lung compliance, or Po₂ levels over time (Fig. 5D).

DISCUSSION

The main findings of the study were: 1) in a severe PA103 bacterial pneumonia in ex vivo perfused human lungs, administration of HMW HA restored and increased AFC rate in a dose-dependent manner. In marginal lungs with baseline AFC < 10%/h (13), administration of HMW HA also numerically reduced both protein permeability and total lung weight gain, indicative of a decrease in pulmonary edema (Fig. 1); 2) the beneficial effects may be partially attributed to suppression of alveolar inflammation and a decrease in the inflammatory properties of plasma EVs (i.e., decrease in inflammatory cytokines). The predominant cellular sources of plasma EVs identified were from platelets and endothelial cells, although the contributions from immune cells cannot be disregarded due to the majority of EVs being exosomes (Figs. 1 and 3); 3) surprisingly, the therapeutic effects of HMW HA were not observed in human lungs with an intact alveolar epithelium and endothelium as seen with an AFC > 10%/h injured with PA103 pneumonia, suggesting that otherwise “healthy” lungs were able to compartmentalize the severe pneumonia with limited pulmonary edema formation (Fig. 5); 4) although in vitro HMA HA coincubation increased phagocytosis of PA103 bacteria by human monocytes and neutrophils, there were no beneficial effects of HMW HA in reducing alveolar bacterial load in both marginal and intact human lungs injured with PA103 pneumonia (Figs. 2, 4, and 5); and 5) lastly, LMW HA had no therapeutic effects in increasing bacterial phagocytosis by human monocytes, suggesting that the MW of the HA was critical for its biological response (Supplemental Fig. S5).

Hyaluronic acid is a nonsulfated glycosaminoglycan in the lung, which is mainly distributed in the peribronchial and inter- and perialveolar space and critical for maintaining homeostasis (14). HMW HA with a molecular weight >1 MDa, composed of repeating polymeric disaccharides d-glucuronic acid and N-acetyl-d-glucosamine, is the main form of HA in the extracellular space. In multiple lung diseases (15, 16), HA undergoes degradation by lysosomal hyaluronidases, reactive oxygen and nitrogen species, and inflammatory mediators (17). The degradation products, low molecular weight HA (< 500 KDa), can decrease endothelial cell barrier function, stimulate angiogenesis, and induce inflammation (18). Surprisingly, based primarily due to its molecular size, HMW HA has the opposite properties of LMW HA; HMW HA exert anti-inflammatory and immunosuppressive effects, whereas LMW HA stimulate gene expression and synthesis of proinflammatory cytokines (19). Previous publications demonstrated that HMW HA administration enhanced endothelial cell barrier properties, reducing protein permeability and the influx of immune cells in LPS or ventilator-induced lung injury (20, 21). Therefore, many investigators have focused on the therapeutic use of exogenous administration of HMW HA in lung diseases. HA influences cell behavior through binding to various cell surface receptors such as CD44, Toll-like receptor (TLR) 2 and 4, hyaluronan-binding protein 2 (HABP2), or receptor for hyaluronan-mediated motility (RHAMM) (22). CD44 is commonly expressed on innate and adaptive immune cells such as macrophages, neutrophils, etc., and EVs released by these immune cells.

Our previous study demonstrated that instillation of HMW HA ameliorated lung injury induced by E. coli bacterial pneumonia and its derived EVs in ex vivo perfused lungs in part by binding to EVs, possibly via a receptor (i.e., CD44) on the EVs, and preventing its interaction with target cells (4). In the current study, IV administration (into the perfusate) of HMW HA also improved AFC and numerically reduced pulmonary edema and inflammation in ex vivo perfused human lungs injured with PA103 pneumonia but only in marginal lungs (Fig. 1). In addition, unlike our previous study, treatment of HMW HA did not reduce the alveolar bacterial CFU load at 6 h in either marginal or intact human lungs (Figs. 2 and 5), although HMW HA treatment increased PA103 bacterial phagocytosis by human monocytes and neutrophils in vitro (Fig. 4). The lack of benefit of HMW HA in human lungs with AFC > 10%/h and in reducing the total alveolar bacterial load in either marginal or healthy human lungs may reflect a lower degree of virulence of PA103 bacteria compared with E. coli bacteria during the early inflammatory response to the pneumonia or too short of a duration of study. In addition, there is a possibility that the antimicrobial effect of HMW HA is bacteria specific; E. coli may be more susceptible to HMW HA than PA103 bacteria. Future studies are planned in exploring the therapeutic effects of HMW HA for longer duration of incubation (i.e., 10 h) and in other clinically relevant pneumonias such as from Streptococcus pneumoniae.

Once considered cellular debris, EVs released into the plasma have now been studied as biomarkers and mediators of cellular communication with both injurious and reparative properties (23–26). For example, plasma EV levels have been considered as a prognosis index for ARDS (27, 28). Several laboratories have also demonstrated that intratracheal or intravenous administration of EVs generated by stimulated endothelial cells or macrophages or collected from the blood of LPS injured rats could induce ALI (29–31). Although there are significant overlaps between EV groups in terms of size and cellular origin, EVs can be broadly classified into exosomes, which originate from intraluminal vesicles; microvesicles, which are generated from outward budding and fission of the plasma membrane; and apoptotic bodies (32). In the current study, both exosomes, with a diameter range of 20–200 nm, and microvesicles, with a diameter range of 200–1,000 nm, were found in the plasma EVs. By flow cytometry, the predominant cellular sources of plasma EVs identified at 6 h were from platelets and endothelial cells (4). EVs derived from monocytes or neutrophils were largely absent, which may be a limitation of using flow cytometry to identify cellular sources (Fig. 3).

EVs are believed to exert their biological effects according to its interaction with target cells through possible transfer of various bioactive molecules including mRNA, microRNA, proteins, and organelles. In the current study, incubation with plasma EVs provoked an inflammatory response in unstimulated human monocytes (i.e., secretion of TNFα by the monocytes), which was suppressed by HMW HA whether administered in vivo during lung perfusion or exogenously in vitro. In addition, in the ex vivo perfused human lung injured with PA103 bacterial pneumonia, administration of HMW HA numerically decreased mRNA content for TNFα and IL-6 in the plasma EVs.

Although there was no effect on total alveolar bacterial load in the ex vivo perfused human lung injured with PA103 bacterial pneumonia, coincubation with HMW HA increased the phagocytosis of PA103 bacteria by LPS-stimulated human monocytes and neutrophils in vitro (Fig. 4). There are several possible explanations for the antimicrobial effect of HMW HA: 1) HMW HA may enhance bacteria phagocytosis by immune cells, possibly through activation of CD44, which is involved in phagocytosis (33, 34); Lee et al. (35) recently found that HMW HA administration significantly increased E. coli bacterial phagocytosis by RAW264.7 cells in part through increased phosphorylation of ezrin/radixin/moesin, a known downstream target of CD44; 2) HMW HA may interfere with bacterial adhesion to a cellular substrate (36), as adhesion to oral and airway cells is necessary for the first step of microbial colonization and pathogenesis; 3) HMW HA appears to have bacteriostatic properties (37, 38); and 4) lastly, the beneficial effect of HA on endothelial barrier properties may prevent translocation of bacteria (39).

There are some limitations to the current study: 1) a short duration of lung injury (i.e., 6 h) that limits the assessment of whether HMW HA can suppress PA103 bacterial growth; 2) lack of exploration of the effects of HMW HA on exotoxins released by Pseudomonas aeruginosa, which are involved in the pathogenesis of lung injury. Unlike PA103, other strains of PA that do not release significant amount of exotoxins may respond differently to HMW HA as a therapeutic (6, 40). In addition, further studies will be needed to address the potential inflammatory role of bacterial vesicles in the pathogenesis of ALI; 3) lack of an intact lymphatic system in our ex vivo perfused human lung for lung interstitial fluid clearance, which may be relevant if AFC is measured over longer periods; 4) lack of other immune organs such as the spleen or liver that may participate in injury and/or repair and is required for HA metabolism (41); and 5) need for more evidence that the mechanisms of action of HMW HA is EV mediated. As suggested by Supplemental Fig. S2, HMW HA may act directly on the alveolar epithelium/endothelium to restore barrier integrity and thus AFC rate, independent of its effect on EVs.

In conclusion, intravenous administration of HMW HA decreased pulmonary edema formation and alveolar inflammation in marginal ex vivo perfused human lungs injured with PA103 bacterial pneumonia. The therapeutic effects may be a direct effect of HMW HA on the injured alveolar epithelium/endothelium as well as via inhibition of the effects of inflammatory EVs on target cells, similar to our previous findings in E. coli bacterial pneumonia. Regardless, this study emphasizes the potential role of EVs in the pathogenesis of ALI, a rapidly evolving area for experimental and laboratory research (42). Based on these findings, further studies of HMW HA are warranted such as in a large animal model of sepsis (i.e., sheep) to overcome the limitations of the ex vivo perfused human lung and for possible translation to clinical trials.

SUPPLEMENTAL DATA

Supplemental Materials and Methods: https://doi.org/10.6084/m9.figshare.15161541.

Supplemental Fig. S1: https://doi.org/10.6084/m9.figshare.13473450.

Supplemental Fig. S2: https://doi.org/10.6084/m9.figshare.15161508.

Supplemental Fig. S3: https://doi.org/10.6084/m9.figshare.15161517.

Supplemental Fig. S4: https://doi.org/10.6084/m9.figshare.15161526.

Supplemental Fig. S5: https://doi.org/10.6084/m9.figshare.15161529.

GRANTS

National Natural Science Foundation of China No.81930058 (H. Qiu), National Institute of Health National Heart, Lung, and Blood Institute Grant Number HL 140026 and HL 134828 (M. A. Matthay) and HL 113022 and HL 148781 (J. W. Lee).

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

X.Z., S.S., and J.-W.L. conceived and designed research; X.Z., S.S., A.L., Y.N., W.H., H.Q., A.S., T.S., and J.-W.L. performed experiments; X.Z., S.S., and J.-W.L. analyzed data; X.Z., S.S., M.A.M., and J.-W.L. interpreted results of experiments; X.Z., S.S., and J.-W.L. prepared figures; X.Z., S.S., and J.-W.L. drafted manuscript; X.Z., S.S., M.A.M., and J.-W.L. edited and revised manuscript; J.-W.L. approved final version of manuscript.