Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2026 Jun 23.
Published in final edited form as: Anesth Analg. 2025 Jun 23;142(2):403–415. doi: 10.1213/ANE.0000000000007545

Protective and immunomodulatory functions of exogenous B cells in experimental hyperoxic lung injury

Dusan Hanidziar 1,*, Liam J Dwyer 2, Sylvia L Ranjeva 1, Eva Csizmadia 3, Saumya Maheshwari 2, Juan D Valencia 3, James W Aspden 2, Nina Farhat 3, Leo E Otterbein 4, Simon C Robson 3, Ruxandra F Sîrbulescu 2,5, Mark C Poznansky 2
PMCID: PMC12906985  NIHMSID: NIHMS2065140  PMID: 40549576

Abstract

Background:

Acute respiratory distress syndrome (ARDS) is a result of diffuse lung injury and dysregulated inflammation. Recent studies have demonstrated that B cells can perform anti-inflammatory and tissue-protective functions. We hypothesized that systemic B cell administration could have therapeutic effects in hyperoxic acute lung injury.

Methods:

Acute lung injury was modeled in adult C57BL/6J male mice through continuous exposure to hyperoxia (FiO2 >90%). Mature B cells (CD45R+/CD19+) were purified from spleens of age- and sex-matched C57BL/6J mice. B cells (107) or saline were administered intravenously after 24 hours of hyperoxia. Hyperoxia exposure was continued for up to 96 hours. The effects of adoptive B cell therapy were assessed using histologic, physiologic (pulse oximetry, echocardiography), and immunologic (flow cytometry) readouts.

Results:

Hyperoxia led to a 50% depletion of endogenous pulmonary B cells by day 3, from 30% to 15% CD45+ lung immune cells (95% CI 6.16 – 24.45; p = 0.0017). B cell administration ameliorated B cell loss, improved lung injury scores (median score in saline-treated = 3.0 vs. B cell-treated = 2.67; p = 0.0101) and lung cellular infiltration (F [2,34] = 11.99; p = 0.0001). By day 3, B cells limited the duration of oxygen desaturations (difference 0.39 s; median length = 1.01 s in saline-treated vs. 0.62 s in B cell-treated; 95% CI 0.02 s – 0.73 s; p = 0.03) and their depth (median nadir = 82.0% in saline-treated vs. 85.9% in B cell-treated, 95% CI −6.6% – −0.84%; p = 0.04). B cell-treated mice showed a median 3.82% increase in left ventricular ejection fraction by day 3, compared to 12.35% in saline-treated mice (mean difference 7.32%; 95% CI −5.0% – 19.6%; p = 0.23). Exogenous B cells represented less than 1.5% of pulmonary B cells on day 3. B cell administration had homeostatic effects on relative abundance of pulmonary immune subsets affected by hyperoxia, including endogenous B cells, CD4+ T cells, NK cells, monocytes/macrophages, and neutrophils. Significant immunomodulatory effects of B cell administration were observed in myeloid cells in the lungs and included reductions in the proportion of IL-17-expressing Ly6Clo monocytes (F [2,14] = 19.02; p = 0.0001), alveolar macrophages (F [2,14] = 10.32; p = 0.0018), and neutrophils (F [2,14] = 6.621; p = 0.0095) as well as IFNγ-expressing Ly6Clo monocytes (F [2,14] = 48.83; p = 0.0001).

Conclusions:

Our data indicate that adoptive B cell therapy ameliorates hyperoxic lung injury and may represent a novel treatment for ARDS.

Introduction

Acute respiratory distress syndrome (ARDS) develops in at least 10% of mechanically ventilated patients in intensive care units. Patients with ARDS commonly receive high fractions of inspired oxygen to ensure adequate oxygenation.1,2 The impact of hyperoxia in patients is not well characterized because multiple superimposed insults (e.g., infection, mechanical ventilation, hyperoxia) can exacerbate pulmonary inflammation.3,4 In experimental animal models, prolonged hyperoxia triggers pulmonary inflammation and acute lung injury resembling ARDS.5–9

We have recently shown that B cells are the most abundant immune cell population in the lungs of adult mice and found that hyperoxia exposure (FiO2>90%) leads to the depletion of pulmonary B cell memory/regulatory subset.10 B cell depletion also occurs in sepsis11,12 and COVID-19 ARDS.13–15 Transcriptomic data from mouse models and human bronchoalveolar lavage samples also show that B cell signatures are decreased during acute lung injury.16

In murine models of pneumonia and sepsis, endogenous B cells suppress neutrophilic inflammation17,18, IL-17/Th17 responses18, and tissue pathology17–19. Notably, high resistance to hyperoxic lung injury is found in leukemia-inhibitory factor (LIF) transgenic mice that develop B cell-rich follicles in the lungs.20 We have demonstrated tissue-protective and immunomodulatory functions of exogenously administered B cells in mouse models of skin wound healing and traumatic brain injury.21,22 Topical application of B cells on a full-thickness excision skin wound is associated with accelerated wound closure and modulation of neutrophil infiltration at the injury site.21 In a mouse model of traumatic brain injury, intraparenchymal B cell administration has significant neuroprotective effects, reducing brain lesion volume and alleviating motor and cognitive deficits.22 Purified mature B cells applied at the sites of tissue injury respond to the inflamed microenvironment in a TLR-dependent manner, producing a complex cytokine signature, which modulates the phenotypes of stromal cells as well as immune cells, thereby promoting tissue repair.23

Based on these documented immunomodulatory and tissue-protective capacities, as well as the apparent sensitivity of B cells to depletion in the context of hyperoxia and ARDS, we hypothesized that intravenous administration of purified mature B cells will attenuate hyperoxic lung injury (primary outcome). Secondary outcomes included hypoxemia, left ventricular ejection fraction (LVEF), frequencies of major immune cell subsets, and cytokine responses by the immune cells. Our results indicate that adoptive B cell therapy decreases the severity of lung injury, improves oxygenation, and promotes immune homeostasis in the lungs.

Methods

Mouse model of hyperoxic lung injury

Eight-week-old male C57BL/6 mice (JAX Stock# 000664) and BL/6 CD45.1 mice (JAX Stock# 002014) were purchased from Jackson Laboratory (Bar Harbor, ME) and housed in an AAALAC-accredited animal research facility. Mice between ages 9 and 12 weeks were used for experiments. Continuous hyperoxia exposure was performed for up to 4 days, as previously described by our group.10,24 The inspired fraction of oxygen was maintained at above 90% by a fresh flow of oxygen at 2–3 L/min through a plexiglass chamber and confirmed with a calibrated oximeter. Mice were briefly removed from the hyperoxia chamber for intravenous injections and physiologic measurements. All the experiments were approved by the Institutional Animal Care and Use Committee (IACUC) of Beth Israel Deaconess Medical Center.

B cell isolation

Donor spleens were dissociated mechanically through a 40 μm cell strainer in EasySep™ Buffer (STEMCELL Technologies, Vancouver, CA). Negative immunomagnetic B cell selection from splenocyte suspension was performed using a commercially available B cell isolation kit (STEMCELL Technologies) as described previously.21–23 Flow cytometry validated the B cell isolation procedure, which yielded >98% mature naїve CD45R+/CD19+ B lymphocytes, with under 1% contamination with other leukocytes.

B cell administration

B cells were administered in C57BL/6 mice after 24 hours of hyperoxia exposure. Purified B lymphocytes were resuspended in sterile PBS at a concentration of 33 × 106 cells/ml. Animals were anesthetized using ketamine (100 mg/kg) /xylazine (10mg/kg), and 300 μl of B cell suspension (10 × 106 B cells) was administered intravenously via the tail vein, using a 25G needle. Control animals received an equivalent volume of sterile PBS.

Non-invasive physiologic monitoring

Oxygen saturation was measured using the MouseOx Plus system (STARR Life Sciences, Oakmont, PA). The collar sensor was applied to the shaved neck of mice, and the mice moved freely during monitoring. After a period of plethysmography signal optimization (1–2 minutes), a continuous recording was initiated at a rate of 15 data points per second. Error coding of the data was enabled. In each mouse, recordings were performed for at least 5 minutes, and time-stamped data were then exported for further analysis. Transthoracic echocardiography (Vevo 2100 Imaging system) was performed on mice under isoflurane anesthesia, and LVEF was quantified using built-in analysis software.

Tissue collection for histology

Mice were deeply anesthetized with intraperitoneal ketamine (100 mg/kg) /xylazine (10 mg/kg), and the abdominal aorta was transected to allow exsanguination. Lungs were removed en bloc from the thoracic cavity and slowly filled with 500 μl of neutral buffered formalin via a 22G cannula inserted into the trachea. This volume represented approximately 50% of total lung capacity and was sufficient to expand the lung lobes without causing perforation in most mice. If a perforation of a lung lobe occurred, the mouse was excluded from further histologic analysis. After the infusion of formalin and ligation of the trachea, the lungs were immersed in formalin for 24–48 hours. Lungs were embedded in paraffin, and 5 μm tissue sections were obtained at appropriate depth to include left lung (LL), right upper lobe (RUL), and right lower lobe (RLL) on the same microscopy slide.

Histologic evaluation of acute lung injury

Lung tissue sections were stained with hematoxylin and eosin (H&E) using a standard protocol.25 The key histopathologic features of hyperoxic injury include enlargement of alveolar spaces due to alveolar septal damage26, vascular congestion26, thickening of alveolo-capillary membranes due to accumulation of edema and blood-borne elements5,27,28, formation of perivascular cavities due to edema28, and eventually alveolar collapse/consolidation. Given the lack of a broadly recommended lung injury score29,30, key investigators (D.H., R.F.S., S.C.R, M.C.P) developed a lung injury scoring template based on the presence of these features (Supplementary Fig. S1). Two investigators (D.H. and R.F.S.) independently assigned the injury scores ranging from 1 to 4 to each of the three evaluated lung lobes (LL, RUL, RLL) (Supplementary Fig. S2). Inter-observer agreement was >90%, and the differences were subsequently adjudicated to determine the final score. The final global injury score for each mouse was calculated by averaging the scores from the individual lobes.

Immunohistochemistry staining

Tissue sections were deparaffinized and sodium citrate buffer (pH 6.0) was used for antigen retrieval. Hydrogen peroxide (H2O2) was applied to tissue sections for 10 minutes and sections were then blocked with 7% horse serum (Vector Laboratories, Burlingame, CA) for 30 minutes. Purified rat anti-mouse CD45R/B220 antibody, clone RA3–6B2 (BD Pharmingen, 553084) was applied on the sections overnight at 4°C. The slides were then incubated with biotinylated rabbit anti-rat IgG antibody (Vector laboratories, BA-4001) for 1 hour at room temperature. Vectastain Elite ABC kit followed by DAB Substrate kit (Vector Laboratories, Burlingame, CA) were used to develop brown staining.

Microscopy and Image Analysis

Images of each studied lung lobe (LL, LUL, RLL) were captured using a Nikon Eclipse E600 microscope (Nikon Instruments, Melville, NY) at 40x, 100x and 200x total magnification. The images aimed to capture continuous areas of lung parenchyma lacking large bronchi or branches of pulmonary artery (Supplementary Fig. S2) to make them suitable for injury scoring and quantitative analyses of cellularity. Image analysis, including cell/nuclei counts was performed using built-in functions in ImageJ software (NIH, Bethesda, MD).

Isolation of immune cells from lung parenchyma and blood

Brefeldin A (250 μg) was administered intravenously in anesthetized mice 4–6 hours before blood and lung harvesting. On day 3 of hyperoxia exposure, mice were deeply anesthetized, and 0.6–1.0 ml of blood was collected via subxiphoid cardiac puncture. After thoracotomy, the lungs were carefully excised avoiding the inclusion of the thymus, mediastinal lymph nodes and central airways. Red blood cells (RBCs) were removed from blood samples by incubating with 10 ml of 1x RBC lysis buffer (eBioscience) for 10 min at room temperature. Lung samples were minced using fine scissors and enzymatically dissociated using Liberase™ TL (Roche). Digested tissues were then passed through a 100 μm filter to yield a single-cell suspension.

Flow cytometry staining

Cell suspensions from lung and blood samples were washed and resuspended in PBS and stained using a Zombie UV fixable viability kit (BioLegend) for 30 minutes at 4°C. The stained cells were then washed and resuspended in PBS containing 1% fetal bovine serum (FBS), 0.01% sodium azide (RICCA Chemical), and 5% FcR blocking reagent (Miltenyi Biotec) for 10 minutes at 4°C. After blocking, cells were incubated with primary surface antibodies for 30 minutes at 4°C. Surface-stained cells were washed and resuspended in fixation buffer (eBioscience) for 20 minutes at 4°C. After centrifugation, the cells were washed in permeabilization buffer (BioLegend) and incubated for 30 minutes at 4°C with the intracellular antibodies. Antibodies are listed in Supplementary Table 1. Labeled cells were analyzed on an LSR Fortessa X-20 flow cytometer (BD Biosciences), and at least 800,000 events were collected from each sample. Data were analyzed using FlowJo software, version 10.8 (TreeStar, Inc., Ashland, OR).

Statistical analyses

The sample size was computed with G*power 3.1.9.7. The numbers of animals for the proposed experiments were based on the assumption that B cells would decrease the mean lung injury score (primary outcome) from 3.2±0.5 (mean, SD) for the hyperoxia/saline group to 2.6±0.5 (mean, SD) for B cell group (decrease by 18.75%) with the effect size=1.2. Using a two-sample, two-sided t-test with alpha=0.05 and power=0.8, the required sample size to evaluate lung injury is 12 per group. To account for the potential non-normal distribution of injury scores, we add 15% more mice per group, with final required sample of 14 mice per group (hyperoxia/saline, hyperoxia/B cell). Immunologic studies utilized 6 mice per group, with sample size informed by our previous studies. Statistical analyses were performed using GraphPad Prism 9 (GraphPad Software). Before statistical testing, datasets were assessed for normal distribution using the D’Agostino & Pearson normality test. For parametrically distributed data, differences among experimental groups were analyzed using unpaired Student’s t-test (for single comparisons) or one-way ANOVA followed by Tukey’s test for multiple comparisons. Data that was not normally distributed (e.g., lung injury scores) were assessed using nonparametric tests, including the Mann Whitney U for group differences or Spearman’s rank correlation. All reported descriptive statistics are estimated marginal means ± standard error of the mean (SEM). P values < 0.05 were considered statistically significant. Pulse oximetry time series were analyzed using the R statistical programming environment (version 4.0.2). Before analysis, data were validated according to error codes provided by the MouseOx Plus system. Data were treated as missing for a given timepoint if the software detected error in the SpO2 value or heart rate value. Missing data was below 25% (Supplementary Fig. S3), and missing data points were inferred using linear interpolation, consistent with other analyses of the SpO2 time series.31 Desaturation (hypoxemic) events were defined as SpO2 ≤ 88%. First-order metrics (number, duration, lowest saturation level) were calculated for each desaturation. The area under the time-saturation curve (AUC), a metric that combines information about the severity (depth) and length of desaturations, was calculated for each unique desaturation event. Between-treatment group comparisons were conducted using a two-sided Wilcoxon rank-sum test. Sample entropy of pulse oximetry time series was calculated for each mouse at baseline (day 0) and throughout hyperoxia exposure (days 1–4). Sample entropy represents the logarithmic probability that a sequence (i.e., sub-time series from the main series) with window length m and degree of tolerance r will be repeated at a later time. The present analysis was conducted using a window length (m) of 2 and tolerance of 0.2 (0.2 x sample standard deviation), consistent with other studies.31 Treatment groups were compared using pair-wise comparisons as well as a mixed linear model that incorporated time as a random effect.

Results

Intravenous administration of exogenous B cells ameliorates hyperoxic lung injury.

To model acute lung injury, we exposed C57BL/6 male mice to continuous hyperoxia (FiO2>90%) for 4 days. After the initial 24 hours of hyperoxia, mice received either a single intravenous dose of 107 syngeneic mature B cells in saline (n=17) or a single dose of saline (n=16) (Supplementary Fig. S4). On day 4 of hyperoxia (92–96 hours), lungs were harvested, and tissue sections were stained to determine lung injury scores, overall cellularity, and frequency of B220+ B cells. Baseline (normoxia) animals (n=4) did not undergo hyperoxia exposure. When compared to baseline (Fig. 1A), the lungs exposed to hyperoxia demonstrated diffuse alveolar damage, thickened interstitial spaces with cellular infiltration, and vascular congestion (Fig. 1B). In B cell recipients, these histologic changes were less pronounced (Fig. 1C), and histologic injury scores were significantly decreased (median injury score in saline-treated = 3.0 vs. B cell-treated = 2.67; p = 0.0101) (Fig. 1D). Injury scores lower than 3 were assigned to 64.7% (11/17) of B cell recipients as compared to 25% (4/16) of the saline-treated mice. We then quantified the lung parenchymal area occupied by hematoxylin-positive cell nuclei, as an indirect measure of cellular infiltration. While hyperoxia significantly increased tissue cellularity in saline-treated mice as compared to baseline, B cell recipients showed similar values of cellularity as normoxia group (Fig. 1E). The analysis of cellular infiltration of lung lobes with different injury scores in saline-treated and B cell-treated mice is shown in Fig. 1F.

Figure 1. Effect of exogenous B cell administration on histological severity of hyperoxic lung injury.

Figure 1.

Open in a new tab

A: Normal lung parenchyma. B: Lung parenchyma imaged on day 4 of hyperoxia exposure in a saline-treated mouse. C: Lung parenchyma imaged on day 4 of hyperoxia exposure in a B cell-treated mouse. Representative images in A, B, and C were obtained from comparable locations in the left lungs, with total magnification of 100x and 200x (insets). Scale bar = 100 μm D: Lung injury scores in saline-treated (n=16), and B cell-treated (n=17) mice. Statistical significance was determined using the Mann-Whitney test. E: Quantitative analysis of lung cellularity expressed as the hematoxylin-positive area of the left lung image obtained at 200x magnification from saline-treated mice in hyperoxia (n=16), and B cell-treated mice in hyperoxia (n=17), or control mice in normoxia (n=4). Statistical significance was determined using 1-way ANOVA followed by the Tukey’s correction for multiple comparisons. F: Cellular infiltration (hematoxylin-positive area) of individual lobes with different histology scores (2, 3, and 4), in saline-treated and B cell-treated mice. Statistical significance was determined using 1-way ANOVA followed by the Šídák correction for multiple comparisons. *p<0.05, **p<0.01, ***p<0.001, ns, not significant.

We next evaluated the impact of B cell administration on the frequency of B cells in the lung parenchyma. We show that B220+ B cells are abundant in the lungs at baseline (Fig. 2A, D), similar to our previous report.10 However, the proportion of B cells relative to all nucleated cells in the lung parenchyma is markedly reduced after 4 days of hyperoxia (Fig. 2B, D). The decrease in B cell proportion is mitigated in B cell recipients (Fig. 2C, D). B cell frequencies in the lung parenchyma showed a non-significant (Spearman r = −0.2387, p = 0.2065) negative overall correlation with lung injury scores (Fig. 2E).

Figure 2. Quantitative histological analysis of B cell population in the lungs.

Figure 2.

Open in a new tab

A: Normal lung parenchyma. B: Hyperoxic lung injury on day 4 in saline-treated mice. C: Hyperoxic lung injury on day 4 in B cell-treated mice. Representative images in A, B, and C were obtained from the left lung, B220 and hematoxylin staining, at a total magnification of 200x. Scale bar = 100 μm D: Percentage of B220+ cells among all hematoxylin+ nucleated cells on images of LL, RUL, and RLL. Comparison of uninjured mice under normoxia (n=3), saline-treated mice in hyperoxia (n=6), and B cell-treated mice in hyperoxia (n=8). Statistical significance was determined using 1-way ANOVA followed by Tukey’s multiple comparisons correction. E: Spearman correlation between lung injury severity score of the left lungs and in situ abundance of B cells on the images of the left lung. ***p<0.001, ****p<0.0001.

Exogenous B cells limit hypoxemia and delay increases in left ventricular ejection fraction.

Impairment of oxygenation is the major consequence of acute lung injury, and one of the key diagnostic criteria of ARDS.1,2 To evaluate the impact of B cell administration on oxygenation, we exposed C57BL/6 mice to continuous hyperoxia for 4 days and administered a single intravenous dose of 107 B cells (n=10) or PBS (n=10) after the initial 24 hours of hyperoxic exposure. Oxygen saturation (SpO2) was measured non-invasively at baseline (day 0) and then daily for 4 days (Supplementary Table 2).

SpO2 recordings demonstrated a complex pattern with recurrent desaturations in both saline-treated and B cell-treated mice (Fig. 3A, B). Desaturations on day 3 appeared more severe in the saline-treated group (Fig. 3A) when compared to the B cell-treated group (Fig. 3B). To compare the severity of desaturations between the groups, we calculated the area under the curve (AUC) over each daily recording interval. While no difference between the groups was found on day 0 (baseline), day 1 (hyperoxia, no i.v. treatment), and day 2 (hyperoxia, 16 hours after i.v. treatment), the log AUC was significantly lower on day 3 (hyperoxia, 40 hours after i.v. treatment) in B cell-treated mice (p=0.031) (Fig. 3C). B cell-treated mice showed shorter duration of oxygen desaturations (SpO2 <88%) (difference 0.39 s; median length = 1.01 s in saline-treated vs. 0.62 s in B cell-treated; 95% CI 0.02 s – 0.73 s; p = 0.03) (Fig. 3D) and higher SpO2 values during desaturation events (median nadir = 82.0% in saline-treated vs. 85.9% in B cell-treated, 95% CI −6.6% – −0.84%; p = 0.04) (Fig. 3E; Supplementary Fig. S5). The SpO2 AUC increased markedly in both groups on day 4 (Fig. 3C). These data indicate that the peak beneficial effect manifested 2 days after exogenous B cell administration.

Figure 3. Analysis of oxygen desaturation events from pulse oximetry recordings.

Figure 3.

Open in a new tab

A: SpO2 recordings in saline-treated mice on day 3 of hyperoxia. B: SpO2 recordings in B cell-treated mice on day 3 of hyperoxia. Traces in A and B were constructed from 5-minute-long recordings after excluding error values. A total of 2760 recorded data points (total duration of 3.07 minutes) are shown for each animal. Dotted lines indicate the cutoff level for desaturation events, selected at 88%. C: Area under the curve (AUC) of desaturation events during 5-minute continuous pulse oximetry recording for saline-treated and B cell-treated mice at pre-injury baseline (day 0), and on days 1–4 of hyperoxic lung injury. SpO2 threshold of 88% was used to conservatively define desaturation events. Results are shown on the log scale. Each data point reflects a unique desaturation event. P values reflect inter-treatment-group differences at the significance level of 0.05. D: Time (in seconds) per desaturation event for saline-treated and B cell-treated mice on day 3 of hyperoxic lung injury. E: Lowest oxygen saturation per desaturation event for saline-treated and B cell-treated mice on day 3 of hyperoxic lung injury.

Complex properties of pulse oximetry waveforms can provide further insight into cardiorespiratory function.31–33 We analyzed sample entropy as a metric of SpO2 time-series complexity. Sample entropy of SpO2 time series has been shown to reflect the integrity of cardiorespiratory control systems in disease states such as COPD and sepsis, with higher entropy indicating more preserved cardiorespiratory regulation.33 At baseline (day 0), there was no difference in sample entropy between saline-treated and B cell-treated mice (p=0.85). SpO2 data pooled from days 1–4 showed significantly higher entropy in B cell-treated mice (p=0.0026) (Fig. 4A). A significant decline in entropy was seen in saline-treated mice between baseline and day 3 (p=0.011), while no difference in entropy was seen between baseline and day 3 in B cell-treated mice (p=0.82) (Fig. 4B; Supplementary Fig. S5). B-cell administration had a significantly positive association with sample entropy of the pulse oximetry time series throughout hyperoxia in a mixed linear model that included time as a random effect (inferred odds ratio 1.03, 95% CI 1.02–1.05).

Figure 4. Sample entropy of pulse oximetry time series during progression of hyperoxic lung injury.

Figure 4.

Open in a new tab

A: Sample entropy of pulse oximetry time series for saline-treated and B cell-treated mice at pre-injury baseline (day 0, left) and throughout hyperoxic lung injury (days 1–4, right). Each data point reflects a single entropy value for a single recording time series. P values reflect inter-treatment-group differences at the significance level of 0.05. B: Sample entropy of pulse oximetry time series recordings by day for saline-treated and B cell-treated mice. Each data point reflects a single entropy value for a single recording time series. P values reflect the differences between pre-injury baseline (day 0) and day 3 for each group at the significance level of 0.05.

Cardiac function parameters were assessed by echocardiography at baseline, day 3, and day 4 of hyperoxia. When compared to baseline, we detected increases in LVEF in saline-treated mice on days 3 (p=0.014) and 4 (p=0.00082) (Supplementary Fig. S6A), indicating a hyperdynamic and/or hypovolemic state. In B cell-treated mice, LVEF was not significantly increased on day 3 (p= 0.55) but showed a significant increase by day 4 (p= 0.01) (Supplementary Fig. S6B).

Adoptively transferred B cells promote immune homeostasis and attenuate specific elements of the inflammatory response induced by hyperoxia.

To study the impact of B cell therapy on host inflammatory responses, we exposed CD45.2+ C57BL/6 mice to continuous hyperoxia and administered a single intravenous dose of 107 CD45.1+ B cells (n=6) or saline (n=6) after the initial 24 hours of hyperoxic exposure. Lungs and blood were harvested on day 3 of hyperoxia (between 66–70 hours of hyperoxic exposure) as well as at the baseline (n=6) and processed for immune cell analysis. The expression of 24 immune-related markers was assessed using multiplexed flow cytometry (Supplementary Fig. S7) with nine major immune populations identified in the lungs (Supplementary Fig. S8) and eight in the blood (Supplementary Fig. S9).

Hyperoxia led to a 50% depletion of endogenous pulmonary CD19+B220+ B cells by day 3, from 30% to 15% lung immune cells (95% CI 6.16 – 24.45; p = 0.0017) (Fig. 5B) and a significant B cell depletion in the blood (Fig 5B’). B cell depletion was mitigated in the lungs (Fig. 5B) as well as in the blood (Fig. 5B’) of exogenous B cell recipients. The exogenous CD45.1+ B cells represented less than 1.5% of all B cells in the lungs (Fig. 5A) and in the blood (Fig. 5A’), indicating that they attenuated depletion of endogenous B cells observed in hyperoxia rather than substantially contributed to total B cell numbers. The relative abundance of lung CD4+ T cells and NK cells was significantly decreased in hyperoxia, and this response was also mitigated in B cell recipients (Fig. 5C, F). In peripheral blood, hyperoxia led to significant decreases in the relative abundance of CD4+ and CD8+ T cells, which were mitigated in B cell recipients (Fig. 5C’, 5D’). Hyperoxia also markedly suppressed IL-10 expression in all studied lymphoid populations in the lungs; however, adoptive B cell therapy had no significant impact on their expression of IL-10 (Supplementary Fig. S10). In contrast to the lungs, hyperoxia did not suppress IL-10 expression in lymphoid populations in the blood (Supplementary Fig. S12).

Figure 5. Effect of exogenous B cell administration on the frequency of major immune populations in the lung parenchyma and peripheral blood.

Figure 5.

Open in a new tab

Flow cytometry analysis of immune cell populations in the lungs and peripheral blood of animals maintained for 72 hours under either normoxia + i.v. saline (n=6), hyperoxia + i.v. saline (n=6), and hyperoxia + i.v. B cell therapy (n=6). A, A’: Exogenous B cells infused intravenously can be detected in both lung parenchyma (A) and peripheral blood (A’) of animals exposed to hyperoxia and in controls maintained in room air (normoxia). B-J: Effect of B cell infusion on the relative proportions of major immune cell populations in lung parenchyma. B’-I’: Effect of B cell infusion on the relative proportions of major immune cell populations in the peripheral blood of the same animals. Treatment groups were compared using one-way ANOVA followed by Tukey’s multiple comparisons test. *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001, ns-not significant. Data from 5 control lungs are shown after excluding one sample for technical reasons during data acquisition.

Adoptive B cell therapy had notable immunomodulatory effects on myeloid immune populations. We found that hyperoxia led to significant increases in the percentage of neutrophils in the lungs (Fig. 5G) and in the blood (Fig. 5G’), and this response was attenuated by B cell therapy (Fig. 5G, G’). Hyperoxia also led to a relative decrease in Ly6Clo non-classical monocytes in the lungs, and this change was mitigated by B cell therapy (Fig. 5J). The proportion of CD11chiSiglec-F+ alveolar macrophages was significantly decreased in B cell recipients when compared to baseline (Fig. 5H).

We further examined the expression of major immunoregulatory molecules by myeloid immune populations (Fig. 6A) and identified several significant effects of adoptive B cell therapy (Fig. 6B–G). Most notably, hyperoxia significantly increased the percentage of IL-17 - expressing pulmonary alveolar macrophages, neutrophils, and non-classical Ly6Clo monocytes. B cell administration decreased the percentage of IL-17-expressing Ly6Clo monocytes (F [2,14] = 19.02, p = .0001), alveolar macrophages (F [2,14] = 10.32, p = .0018) and neutrophils (F [2,14] = 6.621, p = .0095) (Fig. 6C–E). Hyperoxia also significantly increased the percentage of IFNγ- expressing pulmonary Ly6Chi classical (Supplementary Fig. S11) and Ly6Clo non-classical monocytes (Fig. 6F). The IFNγ response was significantly suppressed in non-classical monocytes (F [2,14] = 48.83, p = 0.0001) (Fig 6F). Hyperoxia markedly suppressed IL-10 production in all studied myeloid populations in the lungs, and B cell treatment showed no significant impact on IL-10 (Supplementary Fig. S11). In contrast to the lungs, hyperoxia did not trigger comparable IL-17 or IFNγ increases in circulating monocytes, and only a modest IL-17 increase was seen in circulating neutrophils (Supplementary Fig. S13).

Figure 6. Exogenous B cell administration specifically modulates myeloid immune populations and NKT cell in the lung parenchyma after hyperoxic injury.

Figure 6.

Open in a new tab

A: Modal distribution of fluorescence signal intensity for IFNγ, IL-17, IL-10, PD-L1, CD39, and CD73 indicates differential expression on pulmonary myeloid immune subsets. B: In NKT cells, surface expression of CD39 was maintained at baseline levels in mice treated with B cells. C-E: B cell treatment significantly decreased the expression of the inflammatory cytokine IL-17 in neutrophils (C), alveolar macrophages (D), and Ly6Clo monocytes/macrophages (E) in the lung. F: In Ly6Clo monocytes/macrophages, the expression of IFNγ was significantly decreased by B cell treatment. G) In alveolar macrophages, the expression of the surface regulatory molecule PD-L1 was decreased by B cell treatment. Treatment groups were compared using 1-way ANOVA followed by Tukey’s multiple comparisons correction. *p<0.05, **p<0.01, ****p<0.0001, ns-not significant.

Collectively, our data show that B cell administration promotes the establishment of a complex immune homeostatic state in the lungs in which the effects of hyperoxia on the proportions of major immune subsets (B cells, CD4+ T cells, NK cells, monocyte/macrophages, neutrophils) are attenuated and the production of IL-17 and IFNγ is reduced in multiple myeloid immune populations.

Discussion

Our study advances the understanding of the immunologic and physiologic consequences of hyperoxia in a mouse model and describes a novel immunomodulatory cell therapy. We show that hyperoxia leads to profound depletion of B cells in both the lungs and the blood and that systemic B cell administration ameliorates the observed B cell deficit and hyperoxic lung injury. Intravenous B cell administration to treat acute lung injury and models of ARDS has not been previously studied. Hence, our study is foundational in this regard.

We modeled acute lung injury in male C57BL/6 mice that are highly susceptible to toxic effects of hyperoxia34 and develop severe lung injury by day 4 of continuous exposure.5 Others and we have reported that lung pathology in this model shares similarities with human ARDS and is characterized by diffuse alveolar damage, interstitial and alveolar edema, cellular infiltration, vascular congestion, and thrombosis.4–6 We have now described significant increases in hypoxemic events by day 4 of hyperoxic injury. Given progressive respiratory and hemodynamic decline35,36, this model of lung injury can be viewed as highly stringent when compared to models where mice receive a one-time application of a noxious insult (e.g., hydrochloric acid, LPS instillation) and exhibit spontaneous recovery.29,30

Our strategy was to administer B cells when pulmonary inflammation was present but physiologic abnormalities (hypoxemia, circulatory failure) were not yet severe and limiting tolerance of mice to the experimental protocol (e.g., i.v. injections, general anesthesia, echocardiography). We, therefore, administered B cells intravenously after the initial 24 hours of hyperoxia. At this experimental time point, changes indicative of inflammatory activation, such as increased lung cellularity, increased myeloperoxidase levels, sequestration of neutrophils, and perivascular edema, are already present in the lungs.27,28

We estimate that at the time of administration, the dose of exogenous B cells (10 million/mouse) used in our study represents approximately 5% of circulating B cells.37 We found that 48 hours after the infusion, exogenous B cells represented less than 1.5% B cells in the blood and lungs. It is unknown whether higher doses of B cells would exert stronger tissue-protective effects, and dose modification will need to be tested in future studies.

Our study confirms that B cells are the most abundant immune population in the mouse lungs10 and shows that B cells are significantly depleted in mice with hyperoxic lung injury. We also found that hyperoxia led to decreased frequencies of NK cells and CD4+ T cells, but not of NKT cells in the lungs. NKT cells appear to tolerate hyperoxic environments and their detrimental roles were described in our previous studies.24 Neutrophil increases in the lungs of mice with hyperoxic lung injury have been described previously5,27.

In our study, B cell administration had homeostatic effects on the relative abundance of immune subsets affected by hyperoxia, including endogenous pulmonary B cells, CD4+ T cells, NK cells, monocytes/macrophages, and neutrophils. Although exogenous B cells do not completely restore the normal immune cell repertoire, the lungs and blood of B cell recipients develop an intermediate immune phenotype where the relative abundance of several subsets lies between baseline and injury levels. We also found that B cell therapy was highly effective in suppressing IL-17 responses in alveolar macrophages, neutrophils, and pulmonary monocytes, as well as IFNγ responses in pulmonary monocytes. Although molecular profiling of exogenous B cells was not undertaken in our study, we speculate that exogenous B cells may generate IL-10, extracellular adenosine (via CD39 and CD73), or other mediators that may be at least partially responsible for the observed homeostatic and immunomodulatory effects.

We found that B cell therapy was predominantly associated with decreases in pro-inflammatory factors, such as IL-17 and IFNγ, rather than the maintenance or upregulation of canonical regulatory factors (e.g., IL-10, CD73) on host immune cells. IL-10 expression was markedly decreased across pulmonary lymphoid and myeloid subsets in hyperoxia, and this effect was not prevented by exogenous B cells. Other authors have found that hyperoxic injury in mice can be ameliorated with daily injections of IL-10.38 Similarly, frequencies of CD73-expressing B cells, NK cells, and alveolar macrophages were decreased in hyperoxia, and B cell administration did not restore CD73 expression on these cells. CD73 (ecto-5′-nucleotidase) is an ectoenzyme that converts extracellular AMP to adenosine. Adenosine promotes vascular barrier function in hyperoxia, and the loss of CD73-mediated adenosine generation was shown to exacerbate hyperoxic lung injury.39

Our study has several limitations. We tested B cell administration only in male mice, which are highly susceptible to hyperoxia, and experiments needed to be terminated on day 4 when all mice developed severe disease. Female mice were found to be more resistant to the development of hyperoxic injury than male mice.40 B cell administration in female mice could potentially allow for longer monitoring of B cell effects in vivo.

We only tested B cell administration at an early point (24 hours of hyperoxia). Our future studies will explore B cell administration at later time points to mimic clinical scenarios when cell therapy may be administered in patients with severe ARDS. In our model, we observed a loss of protective impact of B cells on oxygenation and cardiac parameters by day 4. However, B cell administration was associated with decreased lung injury scores, decreased cellularity, and improved B cell numbers on day 4. The observed loss of protective impacts on physiologic variables is not fully explained in our study and may be due to the loss of B cell immunoregulatory functions and/or progressive decrease of numbers in tissues, as well as continuous administration of noxious insult ultimately overwhelming B cell-mediated protection. Serial assessments of exogenous B cell frequencies in the tissues and the evolution of their molecular phenotype in vivo will be the subject of our future studies.

ARDS is a syndrome with high mortality, and therefore, new specific treatments need to be developed for clinical use. Allogeneic mesenchymal stem cells and allogeneic Treg cells have been tested in clinical trials of ARDS, but further research and product development are necessary for their broader clinical adoption. Our data and the data of other authors suggest that harnessing the immunomodulatory functions of B cells may represent an additional and clinically translatable treatment for ARDS.

Supplementary Material

Supplementary Material

Key points.

Question:

Does systemic administration of exogenous B cells improve outcomes in experimental hyperoxic lung injury?

Findings:

Administration of exogenous B cells decreases the severity of hyperoxic lung injury, improves oxygenation, and exerts broad immunomodulatory effects in the lungs.

Meaning:

Adoptive B cell therapy represents an innovative approach to the treatment of acute respiratory distress syndrome (ARDS).

Acknowledgements

SCR, RFS, and MCP contributed equally

Disclosure of funding:

This work was supported by grants from the NIH National Heart, Lung, And Blood Institute (K08HL141694, R03HL171347) awarded to DH; National Institute of Neurological Disorders and Stroke (R01NS117598) awarded to RFS; Vaccine and Immunotherapy Center Innovation Fund to LD, RFS, and MCP; and US Department of Defense Award W81XWH-16-0464 to LEO and SCR.

Footnotes

Declaration of Interest: MCP, RFS, and DH are co-inventors on a patent application on the use of B cells for the treatment of ARDS. SCR is a scientific founder of Purinomia Biotech Inc. and consults for eGenesis, AbbVie, and SynLogic Inc.; his interests are reviewed and managed by HMFP at Beth Israel Deaconess Medical Center; by the conflict-of-interest policies.

References

  • 1.Fan E, et al. ; American Thoracic Society, European Society of Intensive Care Medicine, and Society of Critical Care Medicine. An Official American Thoracic Society/European Society of Intensive Care Medicine/Society of Critical Care Medicine Clinical Practice Guideline: Mechanical Ventilation in Adult Patients with Acute Respiratory Distress Syndrome. Am J Respir Crit Care Med. 2017;195(9):1253–1263. [DOI] [PubMed] [Google Scholar]
  • 2.Madotto F, et al. ; LUNG SAFE Investigators and the ESICM Trials Group. Hyperoxemia and excess oxygen use in early acute respiratory distress syndrome: insights from the LUNG SAFE study. Crit Care. 2020;24(1):125. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Kallet RH, Matthay MA. Hyperoxic acute lung injury. Respir Care. 2013;58(1):123–41. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Hanidziar D, Robson SC. Hyperoxia and modulation of pulmonary vascular and immune responses in COVID-19. Am J Physiol Lung Cell Mol Physiol. 2021;320(1):L12–L16. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Sue RD, et al. CXCR2 is critical to hyperoxia-induced lung injury. J Immunol. 2004;172(6):3860–8. [DOI] [PubMed] [Google Scholar]
  • 6.Amarelle L, et al. Hyperoxia and Lungs: What We Have Learned From Animal Models. Front Med (Lausanne). 2021;8:606678. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Fracica PJ, et al. Responses of baboons to prolonged hyperoxia: physiology and qualitative pathology. J Appl Physiol (1985). 1991;71(6):2352–62. [DOI] [PubMed] [Google Scholar]
  • 8.Matute-Bello G, Frevert CW, Martin TR. Animal models of acute lung injury. Am J Physiol Lung Cell Mol Physiol. 2008;295(3):L379–99. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Jones R, Zapol WM, Reid L. Pulmonary artery remodeling and pulmonary hypertension after exposure to hyperoxia for 7 days. A morphometric and hemodynamic study. Am J Pathol. 1984;117(2):273–85. [PMC free article] [PubMed] [Google Scholar]
  • 10.Hanidziar D, et al. Characterization of pulmonary immune responses to hyperoxia by high-dimensional mass cytometry analyses. Sci Rep. 2020;10(1):4677. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Hotchkiss RS, et al. Sepsis-induced apoptosis causes progressive profound depletion of B and CD4+ T lymphocytes in humans. J Immunol. 2001. Jun 1;166(11):6952–63. [DOI] [PubMed] [Google Scholar]
  • 12.Shankar-Hari M, et al. Activation-Associated Accelerated Apoptosis of Memory B Cells in Critically Ill Patients With Sepsis. Crit Care Med. 2017. May;45(5):875–882 [DOI] [PubMed] [Google Scholar]
  • 13.Anft M, et al. COVID-19-Induced ARDS Is Associated with Decreased Frequency of Activated Memory/Effector T Cells Expressing CD11a+. Mol Ther. 2020;28(12):2691–2702. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Lu J, et al. Documenting the immune response in patients with COVID-19-induced acute respiratory distress syndrome. Front Cell Dev Biol 2023; 11:1207960. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Sosa-Hernández VA, et al. B Cell Subsets as Severity-Associated Signatures in COVID-19 Patients. Front Immunol. 2020. Dec 3;11:611004. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Sweeney TE, Lofgren S, Khatri P, Rogers AJ. Gene Expression Analysis to Assess the Relevance of Rodent Models to Human Lung Injury. Am J Respir Cell Mol Biol. 2017. Aug;57(2):184–192. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Podstawka J, et al. Marginating transitional B cells modulate neutrophils in the lung during inflammation and pneumonia. J Exp Med. 2021;218(9):e20210409. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Kozakiewicz L, et al. B cells regulate neutrophilia during Mycobacterium tuberculosis infection and BCG vaccination by modulating the interleukin-17 response. PLoS Pathog. 2013;9(7):e1003472. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Gawish R, et al. A neutrophil-B-cell axis impacts tissue damage control in a mouse model of intraabdominal bacterial infection via Cxcr4. Elife. 2022;11:e78291. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Wang J, et al. Pulmonary expression of leukemia inhibitory factor induces B cell hyperplasia and confers protection in hyperoxia. J Biol Chem. 2003;278(33):31226–32. [DOI] [PubMed] [Google Scholar]
  • 21.Sîrbulescu RF, et al. Mature B cells accelerate wound healing after acute and chronic diabetic skin lesions. Wound Repair Regen. 2017;25(5):774–791. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Sîrbulescu RF, et al. Intraparenchymal Application of Mature B Lymphocytes Improves Structural and Functional Outcome after Contusion Traumatic Brain Injury. J Neurotrauma. 2019;36(17):2579–2589. [DOI] [PubMed] [Google Scholar]
  • 23.Sîrbulescu RF, et al. B cells support the repair of injured tissues by adopting MyD88-dependent regulatory functions and phenotype. FASEB J. 2021;35(12):e22019. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Nowak-Machen M, et al. Pulmonary natural killer T cells play an essential role in mediating hyperoxic acute lung injury. Am J Respir Cell Mol Biol. 2013;48(5):601–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Cardiff RD, Miller CH, Munn RJ. Manual hematoxylin and eosin staining of mouse tissue sections. Cold Spring Harb Protoc. 2014;2014(6):655–8. [DOI] [PubMed] [Google Scholar]
  • 26.Kumar VHS, et al. Nfib hemizygous mice are protected from hyperoxic lung injury and death. Physiol Rep. 2017;5(16):e13398. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Perkowski S, et al. Dissociation between alveolar transmigration of neutrophils and lung injury in hyperoxia. Am J Physiol Lung Cell Mol Physiol. 2006;291(5):L1050–8. [DOI] [PubMed] [Google Scholar]
  • 28.Yano H, et al. Ultrastructural alteration of pulmonary tissue under conditions of high oxygen concentration. Int J Clin Exp Pathol. 2020;13(12):3004–3012. [PMC free article] [PubMed] [Google Scholar]
  • 29.Matute-Bello G, et al. ; Acute Lung Injury in Animals Study Group. An official American Thoracic Society workshop report: features and measurements of experimental acute lung injury in animals. Am J Respir Cell Mol Biol. 2011;44(5):725–38. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Kulkarni HS, et al. Update on the Features and Measurements of Experimental Acute Lung Injury in Animals: An Official American Thoracic Society Workshop Report. Am J Respir Cell Mol Biol. 2022;66(2):e1–e14. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Bhogal AS, Mani AR. Pattern Analysis of Oxygen Saturation Variability in Healthy Individuals: Entropy of Pulse Oximetry Signals Carries Information about Mean Oxygen Saturation. Front Physiol. 2017;8:555. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Jiang Y, et al. A network physiology approach to oxygen saturation variability during normobaric hypoxia. Exp Physiol. 2021;106(1):151–159. [DOI] [PubMed] [Google Scholar]
  • 33.Gheorghita M, et al. Reduced oxygen saturation entropy is associated with poor prognosis in critically ill patients with sepsis. Physiol Rep. 2022;10(24):e15546. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Whitehead GS, et al. Genetic basis of murine responses to hyperoxia-induced lung injury. Immunogenetics. 2006;58(10):793–804. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Howden R, et al. Cardiac physiologic and genetic predictors of hyperoxia-induced acute lung injury in mice. Am J Respir Cell Mol Biol. 2012;46(4):470–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Peták F, et al. Hyperoxia-induced changes in mouse lung mechanics: forced oscillations vs. barometric plethysmography. J Appl Physiol (1985). 2001;90(6):2221–30. [DOI] [PubMed] [Google Scholar]
  • 37.Roth R, Mamula MJ. Trafficking of adoptively transferred B lymphocytes in B-lymphocyte-deficient mice. J Exp Biol. 1997. Jul;200(Pt 14):2057–62. [DOI] [PubMed] [Google Scholar]
  • 38.Li HD, et al. Exogenous interleukin-10 attenuates hyperoxia-induced acute lung injury in mice. Exp Physiol. 2015;100(3):331–40. [DOI] [PubMed] [Google Scholar]
  • 39.Davies J, et al. Adenosine promotes vascular barrier function in hyperoxic lung injury. Physiol Rep. 2014;2(9):e12155. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Lingappan K, Jiang W, Wang L, Couroucli XI, Barrios R, Moorthy B. Sex-specific differences in hyperoxic lung injury in mice: implications for acute and chronic lung disease in humans. Toxicol Appl Pharmacol. 2013. Oct 15;272(2):281–90. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary Material
NIHMS2065140-supplement-Supplementary_Material.pdf (2.2MB, pdf)

RESOURCES