Live imaging reveals distinct modes of neutrophil and macrophage migration within interstitial tissues

Abstract

Cell motility is required for diverse processes during immunity and inflammation. Classically, leukocyte motility is defined as an amoeboid type of migration, however some leukocytes, like macrophages, also employ a more mesenchymal mode of migration. Here, we sought to characterize the mechanisms that regulate neutrophil and macrophage migration in vivo by using real-time imaging of leukocyte motility within interstitial tissues in zebrafish larvae. Neutrophils displayed a rounded morphology and rapid protease-independent motility, lacked defined paxillin puncta, and had persistent rearward polarization of stable F-actin and the microtubule network. By contrast, macrophages displayed an elongated morphology with reduced speed and increased directional persistence and formed paxillin-containing puncta but had a less-defined polarization of the microtubule and actin networks. We also observed differential effects of protease inhibition, microtubule disruption and ROCK inhibition on the efficiency of neutrophil and macrophage motility. Taken together, our findings suggest that larval zebrafish neutrophils and macrophage display distinct modes of migration within interstitial tissues in vivo.

Keywords: Actin cytoskeleton; Macrophage; Microtubule; Migration; Neutrophil; Zebrafish.

Fig. 1.

Neutrophils and macrophages have different morphologies and motility dynamics in live zebrafish. (A) Schematic diagram of a 3 dpf larva. The blue square corresponds to the imaging area for random leukocyte migration. The orange line corresponds to a tail transection wound to induce directed migration. See Movie 1. (B) Time-lapse images of a neutrophil (left) and a macrophage (right) randomly migrating. Bars are shown on the lower panel to indicate the distance migrated. The white cross marks the initial position of the cell body. Scale bar: 50 µm. (C–F) Macrophage and neutrophil perimeter (C), and roundness (D) were measured. 2D speed (E) and directionality (F) of neutrophils and macrophages was measured by manual tracking of time-lapse videos. Data are representative of three independent experiments. P-values were calculated with an unpaired two-tailed t-test.

Fig. 2.

Macrophages but not neutrophils require proteases for migration and form paxillin puncta. (A) Macrophage random migration is affected when a broad-range protease inhibitor mix is included in the larval medium. Under the same conditions neutrophil migration is not affected. The mean±s.d. speed is shown; P-values were calculated with a least squares means analysis. (B) Transiently expressed paxillin–mCherry forms puncta in macrophages, but has a cytoplasmic distribution in neutrophils. The inset column shows a magnification of the boxed region. Scale bar: 10 µm (2.5 µm for inset). See Movie 2. (C) Paxillin–mCherry puncta are located close to the membrane (zx view and inset) on only one side of the cell (zx and yz projections). Scale bars: 10 µm (z-projections); 20 µm (xy-projection). Asterisks mark a nearby cell not labeled with paxillin. See Movie 3. White arrows in B and C mark the direction of cell movement. (D) Dynamics of paxillin puncta inside macrophage protrusions (20 s intervals). Paxillin puncta disperse before retraction of the protrusion. Arrowheads mark the same paxillin punctum. Scale bar: 2.5 µm. Data are representative of seven independent experiments where all macrophages showed paxillin in puncta (a total of 27/27).

Fig. 3.

Arp2/3 inhibition affects macrophage and neutrophil migration. (A) Time-lapse imaging of neutrophils and macrophages during incubation with CK-666 (200 µM). The inhibitor blocked neutrophil random motility, and strongly impaired macrophage random motility. Mean±s.d. velocity is shown; P-values were calculated with a least squares means analysis. (B) Inhibition of the Arp2/3 complex affects cell morphology. Time-lapse images showing a representative macrophage and neutrophil morphology. Arp2/3 inhibition induced neutrophil rounding (white arrowheads), whereas macrophages generated long and multiple filopodia-like protrusions (white arrows). Scale bar: 25 µm. (C) In response to tail transection, neutrophil recruitment was completely blocked by CK-666, but macrophage recruitment was only partly impaired. Data are mean±s.d. and are representative of three independent experiments; P-values were calculated with a two-tailed unpaired t-test.

Fig. 4.

Microtubule and ROCK inhibition affects both neutrophil and macrophage migration in vivo. (A) EMTB-3xGFP transiently expressed in neutrophils showed microtubules localized at the back of the cell. Very few microtubules were localized at the front (white arrowhead). Macrophages expressing EMTB-3xGFP showed microtubules extending to the back of the cell and into protrusions at the leading edge. Scale bar: 20 µm. Macrophages transiently expressing EMTB-3xGFP showed microtubule loops in some protrusions (magnified views indicated by * and **; white arrows). See Movie 5. Green arrowhead, MTOC. White arrows in main panels indicate direction of migration. Images are representative of three independent experiments. (B) Microtubule inhibition using nocodazole 1 µM increased the velocity of neutrophils but not random macrophage motility. The mean±s.d. velocity is shown; P-values were calculated with a least squares means. Representative images below show that both cell types had morphology changes upon treatment with nocodazole (white arrows). See Movie 6. Scale bar: 50 µm. (C) UrtCH–GFP and Lifeact–mRuby localization in the presence and absence of nocodazole. Nocodazole induced hyperpolarization of UtrCH–GPF in neutrophils, but not macrophages. Actin intensity was quantified and normalized as a percentage of the total. The relative expression of UtrCH–GPF and Lifeact–mRuby is presented in a graph for the front-to-rear axis (bold line indicates trend line of each set of color values). Five cells from three independent experiments were quantified. Scale bars: 20 µm. (D) ROCK inhibition by treatment with 100 µM Rockout blocked macrophage migration, but had only a partial effect on neutrophil migration. Mean±s.d. speed is shown; P-values were calculated with a least squares means analysis. Representative images below the graph show the cell morphology changes for each condition. Scale bar: 50 µm. See Movie 8.