Dual roles for Rac2 in neutrophil motility and active retention in zebrafish hematopoietic tissue

Abstract

Neutrophil homeostasis is essential for host defense. Here we identify dual roles for Rac2 during neutrophil homeostasis using a zebrafish model of primary immune deficiency induced by the human inhibitory Rac2D57N mutation in neutrophils. Noninvasive live imaging of Rac2 morphants or Rac2D57N zebrafish larvae demonstrates an essential role for Rac2 in regulating 3D motility and the polarization of F-actin dynamics and PI(3)K signaling in vivo. Tracking of photolabeled Rac2-deficient neutrophils from hematopoietic tissue also shows increased mobilization into the circulation, indicating that neutrophil mobilization does not require traditionally defined cell motility. Moreover, excessive neutrophil retention in hematopoietic tissue resulting from a constitutively active CXCR4 mutation in zebrafish warts, hypogammaglobulinemia, infections, and myelokathexis (WHIM) syndrome is partially rescued by the inhibitory Rac2 mutation. These findings reveal that Rac2 signaling is necessary for both neutrophil 3D motility and CXCR4-mediated neutrophil retention in hematopoietic tissue, thereby limiting neutrophil mobilization, a critical first step in the innate immune response.

Figure 1. Neutrophil intrinsic Rac2 activity is required for neutrophil wound response

(A) Schematics of Tol2-MPX:mCherry-2A-zRac2 vectors, both wild-type and D57N, injected into wild-type AB zebrafish to generate Tg(mpx:mCherry-2A-Rac2) (Rac2) and Tg (mpx:mCherry-2A-Rac2D57N) (D57N) transgenic lines. (B) Relative expression level of Rac2 mRNA in neutrophils in Rac2 and D57N lines compared with Tg(mpx:mCherry) (mcherry) determined by quantitative RT-PCR. Result is presented as mean±SD (N=3). Ns, non significant, two-tailed paired t test. (C) Sudan black staining of neutrophils at tail wounds at 1 h post wounding (hpw) in 3dpf mCherry, Rac2 or D57N larvae. Scale bar, 50 μm. (D) Quantification of neutrophil wound response shown in (C). n=35 (mCherry), 37 (Rac1) and 42 (D57N). p<0.001, Kruskal-Wallis test followed by Dunn’s Multiple comparison test. (E) Morpholino oligonucleotides (MO)-mediated knock down of Rac2 reduced neutrophil recruitment to tail wound at 2hpw. n=30 (ctrl MO) and 37 (Rac2 MO). p<0.001, two-tailed Mann-Whitney test. (F) Expression of MO resistant mCherry-2A-zRac2WT, but not mCherry alone, in neutrophils rescues defect in wound recruitment in Rac2 morphants. Column 1–2 and 3–4 are siblings, which are used as an internal control for MO efficacy. n=55, 60, 54, 40 for columns 1~4. P<0.001, Kruskal-Wallis test followed by Dunn’s Multiple Comparison test. See also Figure S1, Movie S1.

Figure 2. Rac2D57N expressing zebrafish larvae are more susceptible to bacterial infection

(A) Left, demonstration of zebrafish larvae otic infection. Phenol red tracking dye (pink) was used to visualize bacterial infection. Sudan black staining of neutrophil recruitment 2 h post ear infection (hpi) in Rac2 (middle) and D57N larvae (right) larvae. Scale bar, 50 μm. (B) Neutrophils recruited to the ear 2 hpi were quantified. n=41 (Rac2) and 39 (D57N). p<0.001, two-tailed Mann-Whitney test. (C) Neutrophils recruited to the ear 1 hpi were quantified in control and Rac2 morphants. n=36 (Ctrl MO) and 57 (Rac2 MO). p<0.001, two-tailed Mann-Whitney test. (D) Survival of larvae challenged with ~5 × 10^4 cfu of PAK (pMKB1::mCherry). n=16 (Rac2) and 18 (D57N). (E) Recruitment of leukocytes to otic infection in Rac2 and D57N larvae. Zebrafish larvae from Tg (mpx:dendra2) crossed with Rac2 or D57N lines were infected with PAK (~10^4 cfu) into the left ear. Total leukocyte populations were immunostained with L-Plastin antibody. Neutrophils were visualized by Dendra2 green fluorescence. The right ear of the same fish was shown as control. Images are representative for at least 15 individual fish. Scale bar, 50 μm.

Figure 3. Rac2 activity is required for neutrophil motility and polarity in vivo

(A) Tracking of neutrophil random migration in the head in Rac2 and D57N larvae (lateral view). Scale bar, 50 μm .(B) Quantification of 3D speed of neutrophil random motility in mesenchymal tissues of the head of Rac2 or D57N larvae. n=38 (Rac2) and 45 (D57N). p<0.001, two-tailed Mann-Whitney test. (C) Quantification of 3D speed of neutrophil random motility in mesenchymal tissues in the head of Ctrl and Rac2 morphants. n=25 (Ctrl MO) and 22 (Rac MO). p<0.001, two-tailed Mann-Whitney test. (D, E) Tg(mpx:GFP-PH-Akt) was crossed to Rac2 (D) and D57N (E) lines to visualize the localization of PI(3,4,5)P₃-PI(3,4)₂ in neutrophils randomly migrating in mesenchymal tissues of the head. Ratiometric images were generated using mCherry as a volumetric control. Still images from representative movies of randomly migrating neutrophils in the head are shown. Arrows, pseudopods in (D) and protrusions in (E). (F) Tg(mpx:GFP-UtrCH) was crossed to Rac2 and D57N lines to visualize the localization of stable F-actin. Still images from representative movies of randomly-migrating neutrophils in the head are shown. Upper, arrow, uropod; arrow head, retracting pseudopods. Lower, arrow, protrusion. Scale bar, 5 μm (D–F). (G) Schematics of altered cell polarity in D57N neutrophils. PI(3,4,5)P₃-PI(3,4)₂ primarily localizes to the pseudopods in Rac2 neutrophils. D57N neutrophils form transient protrusions that are enriched for PI(3,4,5)P₃-PI(3,4)₂, but fail to form functional pseudopods. Stable F-actin is polarized and enriched in the uropod in Rac2 neutrophils, but is present in protrusions and is not polarized in D57N neutrophils. See also Figure S2, Movies S2, S3, S4 and S5.

Figure 4. Neutrophilia in Rac2-deficient larvae

(A) Illustration of the location of posterior cardinal vein (PCV) that was imaged for quantification of circulating neutrophils. Neutrophils that circulate through posterior cardinal vein (PCV) per minute were scored. Lower, one still image from the representative movie of Rac2 larvae containing circulating neutrophils within the highlighted PCV (blue lines). Scale bar, 100 μm. (B) Quantification of neutrophils that circulate through PCV in 1 minute (min) in Tg (mpx:mCherry), Rac2 and D57N larvae at 3dpf. n=20 each for mCherry, Rac2 and D57N larvae. p<0.001, Kruskal-Wallis test followed by Dunn’s Multiple Comparison test. (C) Quantification of neutrophils that circulate through PCV per min in control or Rac2 morphants at 2dpf. n=40 each for Ctrl or Rac2 morphants. p<0.001, two-tailed Mann-Whitney test. (D) Sudan black staining of neutrophils in the CHT. (E) Quantification of neutrophil number in the CHT of Rac2 and D57N larvae at 3dpf. n=41 (Rac2) and 30 (D57N). p<0.001, two-tailed Mann-Whitney test. (F) Three-Dimensional volume rendering of neutrophil morphology in the CHT in Rac2 and D57N larvae. Boxed regions are enlarged in lower panel. Founder fish are crossed with Tg (fli1:GFP) to visualize endothelial tissues or stroma. Scale bar, 30 μm. (G) Quantification of cell roundness in the CHT in Rac2 and D57N larvae. n=32 (Rac2) and 23 (D57N). p<0.001, two-tailed Mann-Whitney test. See also Figure S3, Movie S6.

Figure 5. Rac2D57N results in increased release of neutrophils from the CHT

(A) Rac2 or D57N lines were crossed with Tg (mpx:dendra2). Photoactivation converts green fluorescent Dendra2 to red, allowing for fate-mapping 1.5 h post ventral fin wounding. Neutrophils in boxed region (CHT) were photoconverted. Locations of wounds were indicated with the wedged lines. Arrow, a circulating neutrophil. Scale bar, 50 μm. (B) Distribution of photolabeled neutrophils that are mobilized out of the CHT in Rac2 and D57N larvae 2 hpw. Open bar, % of photolabeled neutrophils at wound; Filled bar, % of photolabeled neutrophils in circulation. n=6 larvae for each condition, results are presented with mean±SD. (C) Neutrophils from the CHT in 2dpf D57N larvae spontaneously entered the circulation 18 h after photo-labeling. Still images from representative movies in the head are shown. Location of blood vessel is delineated. Arrow, circulating photoconverted neutrophils. Scale bar, 50 μm. (D) Quantification of (C). Green and red represent total and photolabeled circulating neutrophils, respectively. n=12 each. P<0.001, Kruskal-Wallis test followed by Dunn’s Multiple Comparison test. See also Figure S4, Movie S7.

Figure 6. Rac2D57N partially rescued CXCR4-mediated retention of neutrophils in the CHT

(A) Tg (mpx:CXCR4-WHIM) was crossed with Rac2 or D57N lines and circulating neutrophils were quantified. n=26 each. P<0.001, two-tailed Mann-Whitney test. (B) Fluorescent signal in boxed region was stacked vertically into one-dimensional line at each time point. Kymography of 4 min movies with 2 sec interval is shown. Horizontal lines indicate auto-fluorescent pigments which are stationary throughout the movie. Arrows indicate neutrophils circulating in the posterior cardinal vein (PCV). Scale bar, 100 μm. See also Figure S5.