Swimming Against the Tide: Drugs Drive Neutrophil Reverse Migration

See allHide authors and affiliations

Science Translational Medicine  26 Feb 2014:
Vol. 6, Issue 225, pp. 225fs9
DOI: 10.1126/scitranslmed.3008666


Unbiased screens in transgenic zebrafish reveal drugs that resolve inflammation by driving reverse migration of neutrophils (Robertson et al., this issue).

The resolution of an episode of inflammation is a finely controlled and active process essential for the return of infected or damaged tissue to homeostasis. A key component of resolution is ensuring that recruited inflammatory cells, particularly neutrophils, are cleared from the site of injury. Such clearance avoids inappropriate activation and release of neutrophil antimicrobial-defense mechanisms and their consequent host-tissue injury. Failure of resolution pathways contributes to the excessive and persistent inflammation central to the pathogenesis of multiple human diseases, including asthma, rheumatoid arthritis, and acute respiratory distress syndrome (1). Thus the delineation of successful and endogenously controlled resolution mechanisms offers the potential to develop novel therapeutics for the treatment of human disease.

Mechanisms of neutrophil clearance include local programmed cell death (apoptosis) followed by phagocytosis, removal from inflammatory sites via lymphatic vessels or by entry into the luminal surfaces of organs (for example, mucociliary clearance mechanism in the airways), or retreat from sites of inflammation and their neutrophil-beckoning chemotactic gradients (called “reverse migration”). In this issue of Science Translational Medicine, Robertson et al. describe a transgenic zebrafish model to screen for compounds with proresolution properties and identify pharmacological agents that manipulate neutrophil reverse migration (2).


The directed migration of circulating neutrophils toward a site of inflammation is essential for preservation of host immunity. This movement is mediated initially by primed neutrophils that interact with endothelial cells lining the blood vessels, which is followed by sequential rolling, adhesion, and crawling controlled mainly by interactions with selectins and integrins. Gradients of chemotactic inflammatory mediators guide neutrophils toward preferential sites of migration across the endothelium into the subendothelial space (extravasation); this can occur by either paracellular (between endothelial cells) or transcellular (through endothelial cells) routes, with the paracellular route believed to predominate in vivo (3). After extravasation, neutrophils undergo chemotaxis—directed by chemical gradients—to migrate toward the site of inflammation.

Once at an inflamed site, neutrophils contain and destroy invading pathogens by a combination of phagocytosis and extracellular release of proteases, reactive oxygen species (ROS), antimicrobial peptides, and DNA. Neutrophils also contribute to host-tissue repair, immune-cell activation, and antigen presentation to naïve T cells. Furthermore, neutrophils (i) synthesize and secrete multiple inflammatory mediators to further amplify their own recruitment and that of other leukocytes, (ii) interact with macrophages in order to alter their cytokine production and functional responsiveness, and (iii) participate in transcellular biosynthetic pathways (for example, production of lipid autocoids, some of which have proresolution properties, such as lipoxins).

The direction of neutrophil migration was traditionally presumed to be unidirectional—from bone marrow to circulating blood to inflammatory sites—and migrated neutrophils were thought to die locally by apoptosis when their job was complete. However, neutrophils also have the ability to reverse migrate away from sites of inflammation or chemotactic gradients. Buckley and colleagues (4) identified a subset of human neutrophils that can reverse migrate through a tumor necrosis factor–activated endothelial monolayer in culture (called reverse endothelial transmigration). These authors also observed that a small percentage of neutrophils in blood from patients with systemic inflammation display a phenotype reminiscent of neutrophils that have undergone reverse transendothelial migration, suggesting this phenomenon might be occurring in vivo.

Convincing evidence that neutrophils can reverse migrate in vivo, against the chemotactic gradients that attracted them to an inflammatory site in the first place, was achieved with the use of transgenic zebra­fish embryos. After the induction of sterile tissue injury, recordings of neutrophil trafficking in the transparent tissue of the zebrafish during both the recruitment and resolution phases of the inflammatory response yielded unexpected findings. Most neutrophils that had migrated to the site of tissue injury were subsequently observed to migrate back toward the vasculature in a highly directional fashion (initially termed retrograde chemotaxis), implying an active process in the resolution phase of the response (5). In addition, contemporaneously migrating neutrophils were observed to migrate both toward and away from the wound site, suggesting a cell-specific phenomenon and not just the cessation of migratory signals at later time points in the inflammatory cascade. A subsequent report using transgenic zebrafish that expressed a neutrophil-specific photoactivatable fluorescent protein also reported spontaneous reverse migration of neutrophils and demonstrated that activation of the hypoxia-inducible factor 1α (HIF-1α) pathway led to increased neutrophil retention at inflamed sites, not just by delaying apoptosis but also by reducing neutrophil migration away from the injury site (6).

Evidence of neutrophil reverse migration in vivo has also been observed in mouse models of muscle inflammation with the use of high-resolution four-dimensional confocal imaging of leukocyte transendothelial migration (3). The frequency of reverse migration appeared to be inflammatory-stimulus specific (low frequency with IL-1β–induced inflammation, high frequency with ischemia-reperfusion injury) but not related to the magnitude of the inflammatory insult. The mechanism behind these observations appeared to be dependent on the expression level of junctional adhesion molecule C (JAM-C) at endothelial cell junctions. Human and mouse neutrophils that have undergone reverse endothelial migration in vitro have a high level of intercellular adhesion molecule 1(ICAM-1hi) expression, and this protein serves as a marker of cells that have undergone reverse endothelial migration back into the circulation.

ICAM-1hi neutrophils have been found within the pulmonary vasculature after induction of lower-limb ischemia-reperfusion injury in mice, which has led to the hypothesis that these neutrophils are associated with the systemic spread of inflammation and are, therefore, proinflammatory in nature (3). Indeed, pulmonary inflammation with neutrophil recruitment and alterations in alveolar–capillary barrier permeability were observed as a systemic consequence of lower-limb inflammation. However, the possibility remains that the ICAM-1hi neutrophils within the pulmonary vasculature reverse migrated from within the lungs after induction of inflammation, rather than initially migrating from the original site of inflammation and being the causative factor in the observed pulmonary inflammation. In humans, pulmonary inflammation is a frequent sequela of systemic disease and distant inflammatory insults. Whether depletion of ICAM-1hi neutrophils leads to altered systemic inflammatory responses after an episode of localized inflammation is thus worthy of further study. ICAM-1hi neutrophils show delayed apoptosis and enhanced ROS production and are present in increased numbers in humans with active inflammatory diseases (including rheumatoid arthritis and atherosclerosis) (4), further suggesting that these cells play a role in systemic inflammation.


The new work (2) confirms the utility of zebrafish as a model organism for studies of mammalian immunity (2). The high degree of conservation between the zebrafish and mammalian immune-cell repertoires and the relative ease by which zebrafish can be genetically manipulated and their cellular processes visualized strengthen the translational potential of anti-inflammatory compounds identified in zebrafish screens. Using transgenic zebrafish that express neutrophil-specific green fluorescent protein (7) to visualize the animal’s response to sterile tissue injury, Robertson et al. (2) developed a relatively high-throughput system for performing drug screens using compound libraries. The zebrafish tail-fin injury model has been used to identify compounds that influence apoptosis and, therefore, potentially enhance inflammation resolution (7). The zebrafish model also has confirmed that compounds such as R-roscovitine, shown to enhance inflammation resolution in mammalian systems, mediate their proresolution properties via effects on neutrophil apoptosis (8). The converse is also true: Compounds (for example, wogonin) that induce human-neutrophil apoptosis in vitro and augment inflammation resolution in zebrafish (9) demonstrate efficacy in mouse models of inflammation (10).

Furthermore, the selective timing of drug delivery allows discrimination of compounds that act on neutrophil recruitment versus those that act primarily on proresolution pathways (Fig. 1). This approach permitted Robertson et al. (2) to identify the quinone compound tanshinone IIA as having previously unappreciated proresolution effects. Although the compound accelerated human-neutrophil apoptosis in vitro and zebrafish-neutrophil apoptosis in vivo, this was not the main mechanism behind its in vivo anti-inflammatory effects. Tanshinone IIA administration accelerated migration of neutrophils away from the injury site. Furthermore, modeling of neutrophil migration revealed that the enhanced reverse migration resulted from increased nondirectional migration of neutrophils, rather than the establishment of an active chemorepellent effect. Thus, tanshinone IIA–induced accelerated loss of neutrophils from the injury site occurred mainly as a result of alterations in the migratory pattern of neutrophils rather than changes in neutrophil life span.

Fig. 1 Resolution requires removal.

Neutrophils can be removed from inflamed tissues either by local apoptosis and subsequent phagocytic clearance or by reverse migration. Environmental factors play a role in promoting these events. For example, hypoxia promotes neutrophil retention at inflammatory sites and, along with granulocyte-macrophage colony-stimulating factor, drives neutrophil survival by delaying apoptosis. Tanshinone IIA promotes apoptosis of neutrophils and, to a greater extent, reverse migration of neutrophils from inflamed sites. Neutrophils that have undergone reverse transendothelial migration express high levels of ICAM-1.


The new work raises several intriguing questions. What are the intra- or intercellular molecular events driving tanshinone IIA–induced accelerated reverse migration and does the process occur by the same mechanism as that driving spontaneous reverse migration of neutrophils? Are neutrophils that reverse-migrate during active inflammation different from those that reverse-migrate during inflammation resolution—that is, does the timing of reverse migration during an episode of inflammation alter the eventual message (pro- versus anti-inflammatory) and outcome for the organism? Does interference with reverse migration alter macrophage and adaptive immune cell phenotypes both locally and systemically and, therefore, have additional immune-modulatory properties? The zebrafish used in this study are at a larval stage without a fully developed adaptive immune system. What effects do adaptive immune cells have on reverse migration in vivo? Are neutrophils that migrate and reenter vessels different from those that migrate away from an inflamed site but remain within tissues? It may well be that reverse migration within tissues is a relatively widespread phenomenon, but only in situations with reduced endothelial cell JAM-C expression are the neutrophils able to subsequently undergo reverse transendothelial migration and reenter the circulation. What are the cues that drive a neutrophil to spontaneously reverse the direction of its migration? It is already known that high concentrations of chemotactic factors can lead to G protein–coupled receptor desensitization and chemorepulsion, and these findings may provide a clue to the mechanism in vivo.

Although the potential of tanshinone IIA to affect reverse migration in mammalian systems remains to be tested, previous identification of proresolution pharmacological compounds using zebrafish have subsequently been successfully translated into mammalian models of inflammation (10). However, the big questions remain: What is the function of reverse migration and is it important in physiological settings or in pathological scenarios of human inflammatory diseases?


  1. Competing interests: The authors declare that they have no competing interests.
View Abstract

Stay Connected to Science Translational Medicine

Navigate This Article