PMN recruitment in inflammatory lung injury models follows classical transendothelial migration paradigms requiring PECAM-1 and CD99

Abstract

Immune cells are recruited to sites of inflammation in a stepwise process involving a symphony of signals and receptors. In the systemic circulation, the step at which immune cells migrate out of the blood and across the endothelium, transendothelial migration, occurs via homophilic interactions between leukocyte PECAM-1 and CD99 and endothelial cell PECAM-1 and CD99. Previous work showed that rolling and adhesion of immune cells in the lung vasculature does not follow the classical paradigm of inflammatory recruitment; however, the transmigration step of this process has largely gone understudied. In this study, we demonstrate that polymorphonuclear cells (PMNs) use PECAM-1 and CD99 when transmigrating in response to murine chemical, bacterial, and ischemia/reperfusion lung injury (IRI). We demonstrate that recruitment of PMNs in response to both Gram-positive and Gram-negative bacteria is PECAM-1- and CD99-dependent. We implemented a method of intravital microscopy (IVM) of the pulmonary vasculature after IRI, with which we directly visualized and quantified transmigration. We demonstrate, in real time, that PMN enter the alveoli by crossing alveolar capillaries. Because PMNs are known to be independent mediators of both tissue damage and resolution of inflammation, we tested these effective blocking antibodies for survival effects in models of 50-60% mortality, but found none. In summary, our study shows that the classical transmigration protein interactions are necessary for the transmigration of PMNs into the airspace during response to four distinct inflammatory stimuli.NEW & NOTEWORTHY Previous studies have shown that neutrophil extravasation in the lung was selectin-independent and the requirement for leukocyte integrins was stimulus-dependent. This study demonstrates that PECAM-1 and CD99 are required for PMN transmigration during chemical, bacterial, and ischemia/reperfusion lung inflammation. We show directly in real time, using intravital microscopy, that neutrophils extravasate from alveolar capillaries. Blocking antibodies against PECAM-1 or CD99 prevented transmigration into the lung airspace, just as they prevent transmigration in the systemic circulation.

Keywords: CD99; PECAM-1; polymorphonuclear cells; pulmonary inflammation; transmigration.

Figure 1.. PMN recruitment during inflammatory lung injury.

Inflammation was induced via one of three methods: hydrochloric acid oropharyngeal instillation, S. pneumoniae (Gram-positive) intratracheal instillation, or P. aeruginosa (Gram-negative) oropharyngeal instillation. After 24 hours, bronchoalveolar lavage (BAL) fluid was taken for flow cytometry or tissue sections were taken for imaging. Data represent two separate experiments with dots representing individual mice. Statistics performed by Mann-Whitney test. A. PMN recruitment to BAL fluid 24 hours post bacterial injury. PMNs were gated via flow cytometry on CD45+CD11b+Ly6G+ live cells to discriminate against other recruited inflammatory cells. B. PMN recruitment to BAL fluid 24 hours post acid injury. C. Widefield fluorescence microscopy of control and chemical injury treated LysM-eGFP FVB/n mice 24 hours after inflammatory stimulus. Scale bars = 25 μm. D. Histological sections of murine lung tissue. Scale bars = 50 μm.

Figure 2.. PECAM-1- and CD99- dependence of PMN recruitment after lung injury.

Inflammation was induced and BALs or tissue sections were taken 24 hours later. At the time of inflammatory stimulus, mice were simultaneously injected with a control antibody or a PECAM-1 or CD99 blocking antibody. Data represent three separate experiments with dots representing individual mice. Statistics performed with Mann Whitney test. A. PMN recruitment to lavage fluid after P. aeruginosa injury. B. PMN recruitment to lavage fluid after S. pneumoniae injury. C. PMN recruitment to lavage fluid after chemical injury. D. Confocal microscopy imaging of FVB/n LysM-eGFP mice with blocking antibodies given at time of inflammatory acid stimulus. Asterisks denote alveolar spaces, which contain many PMNs in the control lungs, but many fewer in anti-PECAM or anti-CD99 treated mice. In the mice treated with anti-PECAM-1 or anti-CD99, many PMN appear to be arrested within the alveolar capillaries (arrows). Scale bars = 25 μm.

Figure 3.. Histological imaging of three models of lung injury with and without blocking antibodies.

H&E stained paraffin embedded section of A. chemical injured tissue with non-blocking antibodies administered. B. acid injured tissue with PECAM-blocking antibodies administered. C. acid injured tissue with CD99-blocking antibodies administered. D. P. aeruginosa injured tissue with non- blocking antibodies administered. E. P. aeruginosa injured tissue with PECAM-blocking antibodies administered. F. P. aeruginosa injured tissue with CD99-blocking antibodies administered. G. S. pneumoniae injured tissue with non-blocking antibodies administered. H. S. pneumoniae injured tissue with PECAM-blocking antibodies administered. I. S. pneumoniae injured tissue with CD99-blocking antibodies administered. In untreated, anti-PECAM or anti-CD99 treated acid-instilled lungs, PMN are seen in alveolar capillaries, but reduced numbers are in the alveoli themselves in the anti-PECAM and anti-CD99 conditions. J. Untreated, PBS-instilled control tissue. Blue arrows show fibrin and leukocytes in alveoli; yellow arrows show clear alveoli. Scale bars = 50 μm.

Figure 4.. Presence of PMNs in the vasculature, interstitium, and airspace post-blockade.

Mice were subjected to chemical injury and given an antemortem injection of APC anti-Ly6G retro-orbitally to label circulating PMNs. BALs and left lung tissue were harvested sequentially from each mouse 24 h post-instillation. Data represent two separate experiments with dots representing individual mice. Single cell suspensions were made from the lung tissue and subjected to cytometry. Significance determined by non-parametric ANOVA (Kruskal-Wallis test). A. Flow cytometry gating strategy for the analysis of intravascular, interstitial, and airspace PMNs with representative plots. Numbers in the dot plot refer to the percentage of total cells in that quadrant. B. Using counting beads in the flow cytometry analysis, the number of intravascular and interstitial PMNs present with or without antibody blockade after chemical injury were quantified from murine left lung lobe tissue. C. Quantification of airspace PMNs present with or without antibody blockade after chemical injury quantified from BAL.

Figure 5.. Leukocyte recruitment during warm ischemia-reperfusion injury.

Mice were subjected to warm ischemia-reperfusion injury, with 30 minutes of ischemia followed by 3 hours of reperfusion. Each dot represents a 30-minute video from an individual mouse with two mice imaged per experimental day. Fields were imaged at the same magnification across mice and conditions. A. Timeline of the surgery and imaging. B. Histological sections stained with H&E demonstrating IR injured sections with PMN recruitment at the periphery. C. Number of leukocytes present in the airspace of videos taken 3 h post-reperfusion. Statistical significance determined by Mann-Whitney test. D. Number of leukocytes present in the vasculature of videos taken 3 h post-reperfusion.

Figure 6.. IVM visualization of transmigration during lung ischemia/reperfusion injury.

The white arrows represent the leukocytes in the vessels, followed by green arrows for leukocytes squeezing through the endothelium, and the blue arrows for leukocytes in the airspace. A. Schematic representing the classification of leukocytes as “transmigrating.” Leukocytes were classified as transmigrating if they were initially clearly intravascular, followed by a shape change to move through the endothelium, then either Motion A, spreading out on the alveolar walls, or Motion B, suspended in the alveolus shaking with the effects of breathing motions, as determined by analysis of individual z-planes. B. Representative maximum intensity projection videos of Motion A or Motion B. White arrows show leukocyte transmigrating. Scale bars represent 15 μm. See supplemental videos 1–4. C. Number of leukocytes seen transmigrating. Dots represent individual leukocytes from analysis of three 15-minute videos of individual mice per condition. Statistical tests by Mann-Whitney test. D. Size of the vessel through which leukocytes were seen transmigrating as measured by the dextran label. Dots represent individual leukocytes from analysis of three 15-minute videos of individual mice per condition. Statistical test by Mann-Whitney test.

Figure 7.. Intravital microscopy of ischemia-reperfusion injury.

Mice were subjected to warm IR injury and imaged via live animal imaging on a Nikon multiphoton microscope. Mice were LysM-GFP FVB/n with PMNs/monocytes/macrophages (green) and a non-blocking PECAM-1 endothelial junction label (red). Scale bars represent 50um. Each video was taken for 15 minutes with a z-stack taken every 25 seconds. Arrows are used to show GFP+ cells moving over time. A. Representative images from video of leukocyte recruitment post-IR injury. B. Representative images from video of leukocyte recruitment post-IR injury with simultaneous injection of PECAM-1 blocking antibody. C. Representative images from video of leukocyte recruitment post-IR injury with simultaneous injection of CD99 blocking antibody. Asterisks denote alveolar spaces with multiple neutrophils in A, but far fewer in anti-PECAM or anti-CD99-treated mouse lungs. Also see Supplemental Videos 6–8.

Figure 8.. Quantification of leukocyte behavior after IRI via IVM.

Videos were taken of the subpleural vasculature of the left lobe of the lung 3 hours after resumption of perfusion post-ischemic clamp. Each video represents 15 minutes elapsed time with LysM-eGFP+ PMNs and monocytes/macrophages (green) and PECAM-1 endothelial junctions (red). See supplemental videos 6 – 8. A. Percent of LysM-eGFP+ cells that are stagnant during the 15 minutes of imaging. Each point represents a 15-min video taken from an individual mouse. B. Average distance traveled by leukocytes imaged. 15-minute videos were coded as taking place in hour 1 (0–60 minutes), hour 2 (61–120 minutes), hour 3 (121–180 minutes), or hour 4 (181–240 minutes) post-reperfusion. Transit distance was measured using Nikon NIS Elements software (see methods) and the average distance of all leukocytes from each video was taken. Each condition includes video from at least 2 mice, and the figure represents 12 experimental days. ANOVA statistical test performed.

Figure 9.. Survival of lung injury with blocked transmigration.

Inflammation was induced and mice were given either control antibody or blocking antibody injections daily for 7 days or until deceased. Solid lines indicate survival of control antibody treated mice, while dotted line indicates mice treated with blocking antibodies. All data tested for significance by Log-rank (Mantel-Cox) test. A. Probability of survival of hydrochloric acid injury with or without treatment by PECAM-1 blocking antibodies. Data represents 3 experiments with n=20 mice. B. Probability of survival of hydrochloric acid injury with or without treatment by CD99 blocking antibodies. Data represents 3 separate experiments with n=20 mice C. Probability of survival of P. aeruginosa injury with or without treatment by PECAM-1 blocking antibodies. Data represents 2 experiments with n=22 mice. D. Probability of survival of P. aeruginosa injury with or without treatment by CD99 blocking antibodies. Data represents 2 experiments with n=24 mice.