Leukocyte margination

The distal bronchioles and alveolar spaces are lined by cuboidal epithelium and type I and II alveolar epithelium, respectively, and are fed by the pulmonary circulation. The pulmonary capillary bed comprises many short interconnecting segments, and a capillary pathway from an arteriole to a venule contains approximately 40–100 capillary segments (reviewed in [1,2]). In normal lungs, this bed contains a large pool of marginated neutrophils, monocytes, and lymphocytes, which are thought to be important in host defense. The concentration of neutrophils in the capillary blood measures 20–60 times the concentration in the large systemic vessels. The transit time of neutrophils through this capillary bed is longer, measuring on average 26–40 s, compared with the transit times of plasma or red blood cells of 1.5-4.0 s [1,2,3,4]. This longer transit time is thought to be due to the time required for neutrophils to deform to an oblong shape to pass through the 40-60% pulmonary capillary segments that are narrower than spherical neutrophils [5]. Measurements of neutrophil shape showed that neutrophils in arterioles were nearly spherical (shape factor of 1.1) whereas neutrophils in the capillary bed had shape factors of 1.5 ± 0.3, suggesting that most neutrophils must deform and elongate to travel through the capillary bed [5,6]. These studies indicate that the ability to deform is critical in the traffic of neutrophils through normal pulmonary capillaries.

Leukocyte sequestration

During the acute inflammatory response within regions of the distal lung tissue, neutrophils sequester within the capillary bed. Studies examining the role of known adhesion molecules, including L-selectin, P-selectin, and CD11/CD18, have demonstrated no role for these molecules in the initial stages of sequestration, when neutrophils are stopping and accumulating [7,8]. Once neutrophils have stopped, then L-selectin and CD11/CD18 are required to keep the sequestered neutrophils within the capillaries, at least in response to some stimuli.

In many organs fed by the systemic vasculature, neutrophil emigration occurs through the post-capillary venules. Hemodynamic patterns and selectin-mediated rolling of leukocytes are the major mechanisms through which leukocytes sequester at these sites. However, because rolling does not occur within the pulmonary capillary bed and adhesion molecules seem not to be required [6–8], we and others developed alternative hypotheses that focused on the role of the mechanical properties of neutrophils and other leukocytes, particularly on their ability to deform. Observations from many laboratories demonstrated that inflammatory stimuli, particularly those which bind to the seven transmembrane-spanning G protein linked receptors, induce changes in the cytoskeleton of neutrophils that result in a decreased ability to deform [9,10,11]. These studies led to the hypothesis that stimulus-induced increases in F-actin at the cell periphery led to a decreased deformability, preventing neutrophils from trafficking through the capillary bed and therefore increasing their sequestration at sites of inflammation. The role of these changes in vivo was suggested by observations demonstrating that neutrophils sequestered in the pulmonary capillaries of rabbits after the infusion of complement protein 5 fragments were rounder than neutrophils in the capillaries of control rabbits [9]. Although the hypothesis that this increased stiffness and decreased ability to deform mediates the initial sequestration of neutrophils within the pulmonary capillary bed at sites of inflammation has been difficult to prove directly, a considerable number of circumstantial data have accumulated that support this concept. Current studies are focused on understanding the mechanisms through which inflammatory mediators induce this rearrangement of the cytoskeleton.

The cytoskeletal changes that mediate the decreased deformability seem transient, and adherence between the neutrophils and the endothelial cells occurs within 5–30 min. The mechanical properties of these adherent and crawling neutrophils become very complex. Yanai et al [12] have demonstrated, with the use of optical tweezers, that granules within the cytoplasm of the leading edge are easier to oscillate than granules in the body or the trailing region of neutrophils. Schmid-Schoenbein et al [13] have suggested that pseudopods have more rigid viscoelastic properties than the remainder of the leukocyte. A full understanding of the biochemical and mechanical events that occur during adhesion and crawling has not yet emerged but is an important topic of investigation; the events might be modulated in part through signaling by ligated adhesion molecules. Neutrophil adhesion to endothelial cells induces changes in both neutrophils and endothelial cells that result in cytoskeletal and other intra-cellular alterations that are probably required for the subsequent steps in the response of neutrophils. For example, when neutrophils bind to pulmonary microvascular endothelial cells through intercellular cell-adhesion molecule-1 (ICAM-1), ICAM-1-dependent intracellular signaling responses are initiated [14]. This signaling response seems to require reactive oxygen species generated by xanthine oxidase, and results in the remodeling of the cytoskeleton through signaling pathways that are only now being explained.

Leukocyte emigration

Following sequestration and initial adhesion, neutrophils migrate along the capillary endothelium and into the interstitium and alveoli. There are at least two pathways through which this emigration occurs, one that uses the leukocyte adhesion molecule CD11/CD18 and one that does not. Which adhesion pathway is selected appears to depend upon the stimulus (reviewed in [11]). Stimuli that induce CD18-dependent emigration include Escherichia coli, E. coli lipopolysaccharide, Pseudomonas aeruginosa, IgG immune complexes, interleukin-1, and phorbol myristate acetate. These stimuli seem to act by inducing the translocation of nuclear factor-κB, resulting in the production of inflammatory cytokines (including tumor necrosis factor-α) and ICAM-1 on the pulmonary capillary endothelial cells [11,15]. Stimuli that induce CD11/CD18-independent neutrophil emigration include Streptococcus pneumoniae, group B Streptococcus, Staphylococcus aureus, hydrochloric acid, hyperoxia, and C5a. Although the mechanisms through which the CD11/CD18-independent mechanisms occur are less clearly described, neutrophil emigration occurs in the absence of enhanced ICAM-1 expression on the pulmonary capillary endothelium, and interferon-γ might be an important regulator of this response. This pathway of neutrophil emigration seems not to require known adhesion molecules, because numerous studies blocking the function of these adhesion molecules, either singly or in combination with blocking antibodies, small molecules, or genetic depletion, have demonstrated no defect in CD11/CD18-independent neutrophil migration. Current studies are focused on identifying new genes by using differential display and gene microarrays to compare transcripts made during CD18-dependent and CD18-independent neutrophil emigration.

Neutrophils migrate into the alveolar spaces through a well-defined pathway. The site of sequestration and migration is, as demonstrated by many investigators, the pulmonary capillary bed rather than the post-capillary venules. Neutrophils migrate between endothelial cells through junctions located between the thick and the thin walls of the capillary loops [16]. This junctional site overlies discontinuities in the basal laminae and is the site of interactions with interstitial fibroblasts within the alveolo-capillary walls. The neutrophils then crawl into the interstitium, where they are often in contact with the fibroblasts and may be crawling along them [17]. They then crawl to sites where type I cells are adjacent to type II cells and preferentially migrate between these two cell types to enter the alveolar space. The process of migration is accompanied by an increase in neutrophil volume, from approximately 128 ± 9 μm³ to 266 ± 9 μm³ [18]. Although this increase in volume seems to be mediated through the sodium/proton antiport, the function of the volume change is not yet clear.