|Home | About | Journals | Submit | Contact Us | Français|
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. Non-invasive 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 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.
Neutrophils represent the first line of defense against tissue injury or bacterial infection. In humans, neutrophils are generated in large numbers, up to 2 × 1011 per day, in the bone marrow, and are subsequently released into the circulation, where they comprise approximately 70% of the circulating white blood cells. Upon tissue injury or infection, neutrophils adhere to the endothelium, transmigrate out of the vasculature and subsequently infiltrate into tissues to mediate host defense. Proper generation and distribution of neutrophils within tissue compartments are critical for human health. Reduced neutrophil production or failure to move into the vasculature results in inherited neutropenia syndromes including Warts, Hypogammaglobulinemia, Infections, and Myelokathexis (WHIM) syndrome (Zuelzer, 1964). On the other hand, neutrophil activation can lead to non-specific tissue damage and contribute to the pathogenesis of inflammatory diseases including ischemia-reperfusion injuries, autoimmunity and others (Summers et al., 2010). The vasculature serves as a highway to efficiently transport neutrophils from hematopoietic tissue to sites of tissue injury (McDonald et al., 2010). Therefore, understanding the mechanisms that regulate neutrophil mobilization from hematopoietic tissue into the circulation is fundamental to appreciate innate immune function in health and disease.
Neutrophilia is a hallmark of the normal host response to stress or infection (Summers et al., 2010). Neutrophilia is also seen in Leukocyte Adhesion Deficiency (LAD), a primary immunodeficiency characterized by abnormal neutrophil distribution with increased circulating neutrophils and absent recruitment to tissues or infection (Etzioni and Alon, 2004). There are different types of LAD, according to the underlying genetic deficiency. LAD I, results from partial or complete loss of the β2 integrin CD18, a molecule required for tight adhesion of neutrophils to endothelium, a critical step during neutrophil transmigration. Both neutrophil intrinsic and extrinsic factors may contribute to neutrophilia. Increased circulating neutrophils can result from impaired transmigration and subsequent accumulation in the vasculature, increased survival in the circulation or increased neutrophil production and mobilization from hematopoietic tissue such as with G-CSF exposure (Forlow et al., 2001). Although the CXCR4-SDF1 signaling axis is known to be critical for modulating neutrophil retention in the bone marrow (Summers et al., 2010), the molecular mechanisms that govern neutrophil mobilization, the first step in neutrophil activation, remain poorly understood.
A new type of LAD (LAD IV) is emerging in patients who display LAD-like phenotypes despite normal expression of cell surface adhesion molecules (Pai et al., 2010). An inhibitory mutation in hematopoietic-specific RAC2, D57N has been reported in two infants who presented with recurrent bacterial infections, in combination with neutrophilia (Ambruso et al., 2000; Berthier et al., 2010; Williams et al., 2000). In our recent studies we have shown that neutrophils from patients with the Rac2D57N mutation have impaired polarization and directed migration in vitro (Berthier, 2010). The Rho GTPases Rac1 and Rac2 are key regulators of the actin cytoskeleton and cell signaling (Bokoch, 2005; Filippi et al., 2004). Distinct roles for Rac1 and Rac2 have been identified during neutrophil chemotaxis in vitro. Rac1 is essential for gradient detection and orientation toward the chemoattractant source and mediates uropod retraction whereas Rac2 is the primary regulator of actin assembly, which provides the molecular motor for motility in vitro (Pestonjamasp et al., 2006; Sun et al., 2004). However, it is not known how Rac2 regulates neutrophil polarized migration in vivo. In addition, the role for neutrophil-intrinsic Rac2 function in mediating neutrophil homeostasis in vivo is still under debate. Rac2-deficient mice display mild neutrophilia and defective host defense (Roberts et al., 1999), associated with increased neutrophil production in the bone marrow. Thus, whether Rac2 signaling modulates neutrophil mobilization from hematopoietic tissue has yet to be determined.
Despite its physiological importance, we know very little about the signaling mechanisms that regulate neutrophil mobilization into the vasculature and how this may be altered in disease. Why do patients with LAD and inhibitory Rac2 mutations have increased circulating neutrophils? Is polarized cell migration necessary for neutrophil mobilization from hematopoietic tissue? To address these questions and understand how Rac2 mutations lead to disease, we used zebrafish embryos, which have been an emerging tool for the study of immune responses in live animals using real-time imaging. Here we show that the Rac2D57N mutation in zebrafish neutrophils is sufficient to recapitulate human disease characterized by impaired neutrophil recruitment to infection, peripheral neutrophilia and reduced host survival with P. aeruginosa infection despite normal macrophage responses. Non-invasive live imaging reveals impaired neutrophil polarization and 3D motility with the inhibitory Rac2 mutation. Using photolabeling to track neutrophil fate from hematopoietic tissue, we show increased neutrophil mobilization from the hematopoietic tissue, suggesting that Rac2-mediated motility is not necessary for neutrophil mobilization. Depleting endogenous Rac2 with morpholino oligonucleotides results in similar phenotypes. Moreover, we find that Rac2 signaling is necessary for neutrophil retention and neutropenia in a zebrafish model of WHIM with constitutive CXCR4 signaling. These studies provide insight into the mechanisms that regulate neutrophil homeostasis and demonstrate how altered cell signaling can contribute to the pathogenesis of human immune deficiency.
Zebrafish Rac2 and human RAC2 share 93.8% amino acid homology (Figure S1D) and the residue Asp57 that is mutated in the LAD type IV is conserved. Similar to humans (Heyworth et al., 1994), zebrafish neutrophils express both rac1 and rac2 as determined by RT-PCR (Figure S1E, F), and rac2 is the predominant isoform. Rac3a or rac3b, which are generally more specific to neuronal tissue, are not expressed in zebrafish neutrophils (data not shown). To determine if the Rac2D57N mutation affects Rac2 activity, we ectopically expressed the human Rac2D57N mutation in neutrophil-like HL-60 cells. Expression of Rac2D57N, at a level that does not change the total amount of Rac2 protein, reduces both Rac2 and Rac1 activation in HL-60 cells upon fMLP stimulation (Figure S1A, B), suggesting that Rac2D57N is a potent inhibitor of endogenous Rac activity.
To model LAD type IV in zebrafish, we generated transgenic zebrafish that express either zRac2-D57N (D57N) or zRac2 wild-type (Rac2), together with mCherry linked by a self-cleavable viral 2A peptide under the neutrophil mpx promoter (Mathias et al., 2009; Mathias et al., 2006) (Figure 1A). Using this established multi-cistronic gene expression system, the Rac2 and mCherry are produced as independent proteins from a single transcript due to the inefficiency of peptide bond formation in the viral 2A peptide (Provost et al., 2007; Yoo et al., 2010). The total expression levels of Rac2 in zRac2 wild-type (Rac2) and zRac2-D57N (D57N) are increased approximately 2 and 1.6 fold respectively compared to control neutrophils using quantitative RT-PCR (Figure 1B), indicating that the transgenic lines have minimal increases in total Rac2 expression in neutrophils. Accordingly, ectopic expression of wild-type Rac2 in neutrophils does not significantly alter neutrophil wound responses compared to expression of mCherry alone (Figure 1C, D), suggesting that ectopic expression of wild-type Rac2 does not alter neutrophil function. In contrast, neutrophils in D57N larvae fail to respond to a physical wound at the tailfin (Figure 1C, D and Movie S1), consistent with the impaired in vitro neutrophil chemotaxis reported for LAD type IV patients (Ambruso et al., 2000; Berthier et al., 2010; Williams et al., 2000). The requirement for Rac2 in neutrophil wound recruitment is also seen with morpholino oligonucleotides (MO)-mediated knockdown of Rac2 using a translation blocking MO (Figure 1E). Expression of wild-type Rac2 in neutrophils is sufficient to rescue the impaired wound response in Rac2 morphants (Figure 1F), indicating that Rac2 expression in neutrophils is necessary for wound recruitment.
We next investigated neutrophil recruitment to a localized otic infection with Gram-negative P.aeruginosa. D57N neutrophils fail to respond to P. aeruginosa injected into the left ear of zebrafish larvae (Figure 2A, B). Consistently, neutrophils in Rac2 morphants also do not response to localized infection (Figure 2C). Moreover, D57N larvae are more susceptible to bacterial infection than Rac2 larvae (Figure 2D), consistent with the idea that neutropenic hosts are hypersensitive to P. aeruginosa. Macrophages respond to bacterial otic infection in D57N larvae as indicated by immunostaining for L-Plastin that detects both macrophages and neutrophils in zebrafish larvae (Mathias et al., 2009) (Figure 2E). Macrophages labeled with transient expression of Tol2-mpeg-GFP are also recruited to P.aeruginosa otic infection (data not shown). These findings suggest that macrophages are not sufficient for host defense in zebrafish P. aeruginosa infection. Taken together, ectopic expression of Rac2D57N in neutrophils is sufficient to cause a neutrophil immunodeficiency in transgenic zebrafish, establishing RAC2D57N as the causative mutation in the primary human immunodeficiency LAD type IV.
We next determined how Rac2 regulates neutrophil migration in vivo. Rac1 has been reported to regulate the neutrophil chemotaxis compass and uropod retraction while Rac2 drives random motility in vitro (Pestonjamasp et al., 2006; Sun et al., 2004). In accordance, 3D random motility of neutrophils in the mesenchymal tissues of D57N larvae or Rac2 morphants is completely abolished (Figure 3A, B, C and Movie S2, S3). To determine the molecular mechanism of this motility defect, probes of cell polarization were visualized in the D57N larvae compared to control larvae (Rac2 wild type) in vivo. Polarity is crucial for directed cell migration within both 2D and 3D environments (Yoo et al., 2010). In control neutrophils migrating within tissue (Figure 3C, E and Movies S4, S5), the front of the cell enriches for products of type I phosphoinositide 3-kinases (PI(3)K), PI(3,4,5)P3-PI(3,4)P2, (labeled with PH-Akt-GFP), which in turn activates Rac to mediate actin polymerization and membrane protrusion (Nishikimi et al., 2009). The rear of migrating neutrophils accumulate stable F-actin (labeled with GFP-UtrCH), most likely due to RhoA-actomyosin mediated tail contraction (Yoo et al., 2010). In contrast, in D57N neutrophils, PH-Akt fails to polarize but is noted transiently in small protrusions that do not form functional pseudopods (Figure 3D, E and Movie S4). UtrCH also localizes to these transient protrusions and fails to polarize to the uropod (Figure 3F and Movie S5). Moreover, Rac2 morphants display similar defects in cell polarity and motility (Figure S2). These findings indicate that Rac2 is necessary for the formation of stabilized pseudopods that polarize PI(3,4,5)P3-PI(3,4)P2 and exclude stabilized F-actin (Figure 3G).
We next investigated neutrophil homeostasis in Rac2D57N transgenic zebrafish. Similar to mammals, zebrafish neutrophils are generated during two waves of hematopoiesis. Primitive macrophages that emerge from the rostral blood island can later adopt a neutrophil identity (Le Guyader et al., 2008). Hematopoietic stem cells (HSCs) first give rise to neutrophils at the caudal hematopoietic tissue (CHT) (Murayama et al., 2006), then migrate to the kidney, seeding the site of definitive hematopoiesis in adult zebrafish. In Rac2D57N larvae, neutrophils are distributed in the pericardium region that is continuous with the yolk sac at 3dpf, but not in the head (Figure S3). The pericardium region may correspond to the site where primitive neutrophils emerge at the rostral site, which is otherwise hard to identify when neutrophils actively migrate. Altered neutrophil homeostasis persists at least until 7dpf in the D57N fish (Figure S3). The accumulation of neutrophils in the pericardium region is also observed with Rac2 morphants (data not shown).
Although Rac2-deficient mice are neutrophilic, the requirement of neutrophil-intrinsic Rac activity in marrow retention is still debatable. Neutrophilia in Rac2−/− mice is associated with increased neutrophil presence in the bone marrow (Roberts et al., 1999) that likely results from elevated G-CSF levels and has been suggested to be due to a cell-extrinsic role for Rac2 (Gomez et al., 2008). However, cell-intrinsic Rac activity is required for marrow retention of HSCs (Gu et al., 2003). Therefore, whether neutrophil-specific expression of Rac2D57N is sufficient to induce neutrophilia is difficult to predict. We observe a 10-fold increase in circulating neutrophils in D57N larvae (Figure 4A, B), suggesting a cell-autonomous role for Rac2 in regulating neutrophil mobilization. In contrast, neutrophil expression of WT Rac2 does not affect the number of circulating neutrophils compared with those expressing mCherry alone (Figure 4A, B). The observed neutrophilia does not result from increased neutrophil production since D57N larvae have fewer neutrophils in the CHT (Figure 4D, E). Moreover, there is no detectable difference in G-CSF or G-CSF-receptor expression by RT-PCR in Rac2 and D57N larvae (data not shown). Consistently, Rac2 morphants are also neutrophilic (Figure 4C) but have similar neutrophil numbers in the CHT (data not shown). Neutrophils from D57N fish display an altered, rounded morphology in the CHT, in contrast to the well-spread, polarized morphology in the CHT of Rac2 larvae, and are not motile like control neutrophils (Figure 4F, G and movie S6; data not shown). This morphology is in line with the in vitro finding that Rac2-deficient neutrophils fail to spread on CD18 coated surfaces (Gu et al., 2003; Roberts et al., 1999). Taken together, our findings indicate that Rac2-mediated motility and polarity is dispensable for neutrophil mobilization from hematopoietic tissue in zebrafish larvae.
In Rac2 null mice or patients with LAD, the cause of peripheral neutrophilia is not known. Some studies suggest that neutrophilia may result from increased neutrophil production (Gomez et al., 2008; Roberts et al., 1999), enhanced release of neutrophils into the circulation (Gomez et al., 2008) or alternatively reduced egress from the blood (Roberts et al., 1999). We find that inhibiting Rac2 signaling in a neutrophil specific manner in D57N larvae does not lead to increased neutrophil production (Figure 4D, E) but abolishes neutrophil egress from the blood (Movie S1). To determine if neutrophil mobilization is also altered in D57N larvae, neutrophil fate was tracked in transgenic fish using Tg (mpx:Dendra2), which expresses the photo-convertible protein Dendra2 in neutrophils (Yoo and Huttenlocher, 2011). Photolabeled neutrophils in the CHT migrate to the wound at the ventral fin 1.5 h post wounding in Rac2 larvae, suggesting that expression of Dendra2, or the photolabeling process, does not interfere with normal neutrophil function. However, in D57N larvae, neutrophils do not respond to the wound, but instead, enter the circulation (Figure 5A, B). Even in the absence of wounding, increased numbers of photolabeled neutrophils spontaneously mobilize into the circulation in D57N larvae 18 h post photolabeling (Figure 5C, D and Movie S7). Increased neutrophil mobilization from hematopoietic tissue is also observed with Rac2 morphants (Figure S4). The findings suggest that increased release of neutrophils from the CHT contributes to neutrophilia associated with Rac2 deficiency. In addition, the defect in neutrophil extravasation likely further contributes to the increased number of circulating neutrophils in Rac2-deficient larvae.
The interaction between the chemokine, SDF-1 and its receptor, CXCR4 is a well-characterized signaling axis that mediates bone marrow retention of neutrophils (Summers et al., 2010; Zuelzer, 1964). A transgenic fish expressing an internalization defective CXCR4, Tg(mpx:CXCR4-WHIM-GFP) was previously generated in our lab to model WHIM syndrome (Walters et al., 2010). In the zebrafish WHIM model, constitutive CXCR4 signaling results in neutropenia and abnormal clustering of neutrophils at the ventral side of the head where SDF-1 is expressed (Walters et al., 2010). Neutrophil clustering in rostral-ventral sites where SDF-1 is expressed is partially inhibited when Rac2D57N is expressed in neutrophils in the WHIM transgenic (Figure S5), consistent with the requirement for Rac2 in CXCR4-mediated chemotaxis (Gu et al., 2003). Interestingly, Rac2D57N induces circulating neutrophils in the WHIM transgenic, indicating that the D57N mutation partially rescues CXCR4-WHIM-induced neutropenia (Figure 6A, B). Taken together, these findings suggest a role for Rac2 in CXCR4-mediated neutrophil retention within hematopoietic tissue.
Here we use real-time imaging and photolabeling to investigate how neutrophil fate is altered during development in an inherited immune deficiency in live animals. Using transgenic zebrafish that express Rac2D57N in neutrophils or Rac2 morphants, we demonstrate that neutrophil intrinsic Rac2 signaling is necessary for neutrophil polarization and motility in vivo. We also show that a zebrafish model of LAD, with impaired neutrophil recruitment, but normal macrophage infiltration, is hypersensitive to P. aeruginosa infection. Moreover, we identify a signaling axis through CXCR4 and Rac2 that mediates neutrophil retention in hematopoietic tissue. Taken together, our study provides a perspective on the study of neutrophil homeostasis and how it may be altered or targeted in human disease.
We provide a step-wise dissection of neutrophil responses to localized wounding or bacterial infection with P. aeruginosa. We show that neutrophil infiltration from the vasculature into the site of infection requires Rac2 signaling. Therefore, the zebrafish LAD model recapitulates a hallmark of human LAD characterized by impaired neutrophil infiltration and altered host defense. Interestingly, macrophages are recruited normally in the D57N larvae, but are not sufficient to improve host survival with P. aeruginosa infection. Our finding is in line with previous observations in a murine intrapulmonary P. aeruginosa infection model showing that neutrophil depletion leads to 100% mortality by 2 days post infection (Tsai et al., 2000). This suggests that zebrafish larval neutrophils function similarly to human neutrophils, since human neutropenic hosts are hypersensitive to P. aeruginosa despite normal macrophage responses (Sadikot et al., 2005).
Cell polarization is crucial for cell migration in either 2D or 3D environments (Devreotes and Janetopoulos, 2003; Insall and Machesky, 2009; Kay et al., 2008; Rericha and Parent, 2008; Stephens et al., 2008; Van Haastert and Veltman, 2007; Yoo et al., 2010). Our previous studies demonstrated that spatial-temporal control of Rac activation using a photoactivatable Rac is sufficient to direct cell motility in vivo and induce the polarity of F-actin dynamics and PI(3)K signaling (Yoo et al., 2010). How Rac2 deficiency affects cell polarity in vivo has not been previously characterized. We now show that endogenous Rac2 is necessary for neutrophil polarized cell signaling and motility within tissues in vivo. We observe impaired polarization of PI(3)K products and F-actin dynamics in Rac2-deficient neutrophils. Transient enrichment of the products of PI(3)K is noted at unstable protrusions in Rac2-deficient neutrophils, which is in line with the model that PI(3)K can act upstream of Rac, presumably by recruiting Rac guanine nucleotide exchange factors (Nishikimi et al., 2009). Our results are also consistent with the in vitro observations that Rac signaling is required for a positive feedback loop that leads to full PI3K activation, since the PH-Akt probe is not fully polarized in Rac2-deficient neutrophils (Srinivasan et al., 2003) and localized Rac activation is sufficient to polarize PH-Akt in vivo (Yoo et al., 2010). Surprisingly, Rac2 inhibition leads to a total loss of polarity of the actin cytoskeleton, even in the presence of transient accumulation of PI(3)K products within small protrusions, suggesting that Rac2 signaling is required to regulate the polarity of F-actin dynamics in vivo. Our previous studies demonstrated that Rac-mediated membrane protrusion was not sufficient to remove UtrCH from the front in PI(3)K-inhibited cell in 3D in vivo (Yoo et al., 2010), suggesting that amplification of polarized asymmetry of PI(3,4,5)P3-PI(3,4)P2 by Rac (Srinivasan et al., 2003) may also be necessary for polarizing F-actin dynamics in vivo. Alternatively, it has been shown in vitro that Rac can inhibit RhoA at the cell front through activation of p190RhoGAP (Nimnual et al., 2003) and inactivation of the Rho exchange factor NET1 (Alberts et al., 2005). Although no significant change in total RhoA activity is detected within HL-60 cells lines stably expressing Rac2D57N (data not shown), localized inhibition of RhoA activity may still be impaired in Rac2 deficient cells. In addition, we find significant inhibition of Erk1/2 phosphorylation in Rac2D57N HL-60 cells treated with fMLP (Figure S1C), which is consistent with the observation that Erk1/2 signaling is impaired in Rac2 deficient murine neutrophils (Gu et al., 2003). Whether Erk1/2 signaling is required for neutrophil F-actin polarity and motility in vivo and contributes to the phenotype in Rac2-deficient neutrophils will be a focus of future investigation.
Proper release of neutrophils from hematopoietic tissue is necessary for host defense. In contrast to neutrophil infiltration into tissues, we find that mobilization of neutrophils from hematopoietic tissue into the circulation does not require Rac2. Instead, our results demonstrate a neutrophil-intrinsic role for Rac2, downstream of CXCR4, in active retention of neutrophils in hematopoietic tissue. SDF1 mediated activation of CXCR4 is known to regulate chemotaxis of diverse cell types including primordial germ cells during development, leukocytes in inflammatory responses and the invasive movement of cancer cells in tumor metastasis (Alkhatib and Berger, 2007; Furze and Rankin, 2008; Gelmini et al., 2008; Raz and Mahabaleshwar, 2009). In contrast, in a zebrafish model of WHIM syndrome, we recently demonstrated that SDF1 activation of CXCR4WHIM mediates neutrophil retention in the CHT, suggesting that SDF1 activation of CXCR4 can also restrict neutrophil motility. We now show that the inhibitory Rac2D57N mutation partially rescues neutrophil retention in the WHIM transgenic, indicating that CXCR4-Rac2 signaling induces neutrophil retention. It is known that polarized Rac signaling is sufficient to guide neutrophil motility in vivo (Yoo et al., 2010). However, our findings now suggest that active Rac2 signaling can also mediate static behavior and limit mobilization. It should also be noted that Rac2D57N only partially restores circulating neutrophil numbers in CXCRWHIM zebrafish compared to Rac2D57N fish, suggesting the presence of additional signaling pathways downstream of CXCR4 that can mediate neutrophil retention. Alternatively, expressing Rac2D57N may not result in a complete ablation of Rac2 activity. These two possibilities can be separated when a Rac2 knockout line is generated. Compared with the partial effect with Rac2D57N, Rac2 depletion alone does not rescue CXCR4WHIM mediated neutropenia (data not shown), suggesting that Rac2 morpholino is less efficient at inhibiting Rac2 activity than Rac2D57N. We cannot rule out the possibility that Rac1 may also play a role in CXCR4 WHIM mediated neutrophil retention in hematopoietic tissue. The specific role of Rac1 cannot be tested in zebrafish at this stage because Rac1 depletion results in an early embryonic developmental phenotype (Srinivas et al., 2007). To further address a role for Rac1 in zebrafish neutrophils we generated a transgenic line that expresses the inhibitory Rac1 mutation Rac1T17N in neutrophils (Zhang et al., 2009) and did not observe any significant changes in neutrophil wound response or mobilization (data not shown), suggesting that Rac2 but not Rac1 is critical for neutrophil motility and retention in zebrafish. Taken together, our findings demonstrate that under normal conditions, Rac2 signaling is sufficient to mediate neutrophil retention within hematopoietic tissue.
Human patients with the inhibitory Rac2 mutation (Accetta et al., 2010), D57N, demonstrate peripheral blood neutrophila. Our findings in zebrafish provide a possible explanation for this neutrophilia, at least, in part, through increased mobilization of neutrophils from hematopoietic tissue. The cytoarchitecture of zebrafish CHT is strikingly similar to the bone marrow of mammals (Murayama et al., 2006). Migration across the endothelial barrier is believed to be required for neutrophil egress from hematopoietic compartments (Summers et al., 2010), although it is possible that some neutrophils may develop within the endothelial lumen in zebrafish (Murayama et al., 2006). It is intriguing that Rac2, an essential component for 3D cell motility, is dispensable for neutrophil mobilization, but instead plays an active role in retention. One possibility is that mobilization of neutrophils takes a unique form of cell movement that is independent of Rac2 signaling. It was previously shown that neutrophils mobilize by transmigrating through endothelial cells (transcellular migration) (Burdon et al., 2008). This unique form of transmigration may be independent of Rac signaling. For example, T cells that lack Tiam1, a Rac guanine nucleotide exchange factor, preferentially undergo transcellular migration (Gerard et al., 2009). However, it is also possible that neutrophil mobilization is a passive process that is independent of neutrophil polarity and motility. This question will remain a challenge for future investigation.
In summary, by expressing an inhibitory Rac2 mutation in zebrafish, we established Rac2D57N as the causative mutation in a developmental human immunodeficiency disorder. We have shown that Rac2 is critical for 3D motility and neutrophil polarization of F-actin dynamics and PI(3)K signaling in vivo. We have also demonstrated that Rac2 plays opposing roles in two critical steps during neutrophil homeostasis induced by tissue injury or infection: Rac2 is inhibitory to neutrophil mobilization from hematopoietic tissue but is necessary for infiltration into injured or infected tissues. Finally, our results provide unique insights into the role of Rac2 signaling in active neutrophil retention in hematopoietic tissue, a fundamental process controlling host defense or inflammation, which potentially will aid therapies designed to treat human neutrophil disorders. It is intriguing to speculate that modulation of Rac2 signaling may provide an alternative approach to increase neutrophil mobilization in neutropenic hosts.
HL-60 cells expressing a control plasmid, Rac2 WT, or Rac2 D57N (Berthier et al., 2010) were maintained in RPMI1640 supplemented with 10% fetal bovine serum, 100 U/mL penicillin, and 100 μg/mL streptomycin, at a concentration of 0.1–1 × 106 cells/mL. To differentiate, 1.25% DMSO was added to 2 × 105 cells/ml for 6 days. Differentiation was confirmed through FACS analysis of CD11b expression.
Rho GTPase pulldown assays were performed as described (Lokuta et al., 2003). Briefly, differentiated HL-60 cells were serum starved in mHBSS (150 mM NaCl2, 4 mM KCL, 1 mM MgCl2, 10 mM Glucose, 20 mM HEPES, pH 7.4, 0.2% human serum albumin) and plated on fibrinogen (10 μg/mL) for 5 minutes then stimulated with 1 μM fMLP for 5 minutes and immediately placed on ice and lysed in RIPA extraction buffer (20 mM Tris-HCL, pH 7.5, 150 mM NaCl2, 2 mM EDTA, 2 mM EGTA, 1% TX-100, 0.5% NP-40, 0.25% DOC, 1 μg/mL pepstatin A, 2 μg/mL aprotinin, 1 μg/mL leupeptin, 200 nM phenylmethanesulphonyl fluoride). Forty μg of bacterially expressed GST-PAK-PBD were incubated with the lysates for 1 hour at 4°C. The affinity-precipitated products were run on an SDS-PAGE gel, transferred to nitrocellulose, and immunoblotted for active Rac1 or Rac2 using mouse anti Rac1 (BD Bioscience), rabbit anti Rac2 (Millipore) antibodies. Vinculin is immunoblotted as loading control using mouse anti-Vinculin (Sigma). Whole cell lysate of fMLP activated HL-60 cells were probed for p-Erk1/2 (T185) (Abcam) and pan Erk1/2 (Invitrogen) antibodies.
All protocols using zebrafish in this study were approved by the University of Wisconsin-Madison Research Animal Resources Center. The zebrafish rac2 gene was a generous gift from Dr. Enrique Salas-Vidal (Leiden University, Leiden, The Netherlands). DNA encoding mCherry-2A-Rac2 WT or D57N was PCR amplified and inserted into backbone vector containing minimum tol2 elements, mpx promoter and SV40 polyA sequence. Viral 2A peptide allows the expression of multiple proteins from one transgene in zebrafish (Provost et al., 2007). More than 3 founders (F0) for each construct were generated as described (Walters et al., 2010; Yoo et al., 2010). Experiments were performed with F2 embryos produced by F1 fish. Key experiments were repeated with F2 fish from different founders.
Dissociated cells of Tg (mpx:Dendra2) at 3 dpf were sorted by Fluorescence Activated Cell Sorting (FACS) as described (Walters et al., 2010). RNA was extracted using RNeasy Mini Kit (Qiagen). RT-PCR was performed using 1-step RT PCR Kit (Qiangen) with described primers for mpx, c-fms, ef1a (Dodd et al., 2009) or using the following primers:
Disassociated cells of Tg(mpx:mCherry), Tg(mpx:mCherry-2a-Rac2) and Tg(mpx:mCherry-2a-Rac2D57N) at 3 dpf were sorted by FACS as described (Walters et al., 2010). RNA was extracted using RNeasy Mini Kit (Qiagen). First strand cDNA was then prepared using the SuperSciprt III First-Strand Synthesis System (Invitrogen) with 4 ng RNA as template. cDNA was diluted 1:20 for each PCR reaction. Quantitative RT-PCR (qPCR) was done in triplicate for both the reference, lysozyme C, (forward: TTGTGGTTTAAGCTGTTTGTTGAC; reverse: GGACTGATGGAATAATTTGGAGAC) and rac2 (forward: ACTCTCCTACCC GCAGACG; reverse: CACCTCTGGGTACCACTTGGC) primers. Primers were checked to have similar efficiency and fold change for each line was determined using the delta-delta Ct method. Data were normalized to mpx:mCherry cells.
Tailfin wounding, Immunofluorescent staining and Sudan black staining were performed with larvae at 3 dpf as described (Walters et al., 2010). Bacterial otic infection were performed as described (Levraud et al., 2008). P.aeruginosa strain PAK and PAK (pMKB1::mCherry) were generous gifts from Dr. Dara W. Frank (Medical College of Wisconsin, Milwaukee, WI, USA) and Dr. Samuel M. Moskowitz (University of Washington, Seattle, WA, USA), respectively.
One nl of standard control or Rac2 MO (Gene Tools, 5′-CCACCACACACTTT ATTGCTTGCAT-3′) at a final concentration of 100 μM was injected into the yolk of 1 cell-stage embryos.
Time-lapse fluorescence images were acquired with a confocal microscope (FluoView FV1000, Olympus) using a NA 0.75/20x objective or Nikon SMZ-1500 zoom microscope (Nikon) as described (Yoo et al., 2010). For photoconversion of cells in Tg(mpx:mCherry-2A-Rac2 or D57N, mpx:Dendra2) in the CHT, 405 nm laser was focused into a rectangular region for 3 min with 100% power at 10.0 μ/pixel. Converted Dendra2 protein emits much stronger red fluorescence compared with mCherry. Acquisition settings were adjusted to exclude mCherry signals in non-photoconverted region, allowing tracking of photolabeled neutrophils. For photocenversion of cells in the CHT in Rac2 morphants, 405 nm laser was focused into a rectangular region for 1 min with 70% power at 10.0 μ/pixel. Tracking of neutrophils (Manual Tracking plugin) and quantification of neutrophil shape (Shape Discriptor1u) were performed using ImageJ (NIH, Bethesda, MD).
We would like to thank E. Salas-Vidal, D. Frank and S. Moskowitz for providing us with zrac2 gene and P.aeruginosa strains, respectively. We would like to thank Andrew Bent and laboratory members for critical reading of the manuscript. This work was supported by the National Institutes of Health grant GM074827 (A.H.) and a grant from the Burroughs Wellcome fund (A.H.).
The authors declare no competing financial interests.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.