The family with sequence similarity 3 (FAM3) gene family is a cytokine-like gene family with four members FAM3A, FAM3B, FAM3C and FAM3D. In this study, we found that FAM3D strongly chemoattracted human peripheral blood neutrophils and monocytes. To identify the FAM3D receptor, we used chemotaxis, receptor internalization, Ca2+ flux and radioligand-binding assays in FAM3D-stimulated HEK293 cells that transiently expressed formyl peptide receptor (FPR)1 or FPR2 to show that FAM3D was a high affinity ligand of these receptors, both of which were highly expressed on the surface of neutrophils, and monocytes and macrophages. After being injected into the mouse peritoneal cavity, FAM3D chemoattracted CD11b+ Ly6G+ neutrophils in a short time. In response to FAM3D stimulation, phosphorylated ERK1/2 and phosphorylated p38 MAPK family proteins were upregulated in the mouse neutrophils, and this increase was inhibited upon treatment with an inhibitor of FPR1 or FPR2. FAM3D has been reported to be constitutively expressed in the gastrointestinal tract. We found that FAM3D expression increased significantly during colitis induced by dextran sulfate sodium. Taken together, we propose that FAM3D plays a role in gastrointestinal homeostasis and inflammation through its receptors FPR1 and FPR2.
Chemokines comprise a large family of structurally homologous small cytokines. The classification of chemokines is based on the distance between the first two of the four to six conserved cysteine residues. Four subfamilies of chemokines have been discovered to date: CXC (α), CC (β), C (γ) and CX3C (Zlotnik and Yoshie, 2012). Chemokines can mediate their activities through G-protein-coupled receptors (GPCRs), which have a characteristic seven-transmembrane structure and transduce their signals to the inside of the cell through heterotrimeric G proteins (Rollins, 1997) that are grouped into two families depending on their sensitivity to pertussis toxin (PTX). The Gi/o family is sensitive to PTX, whereas the Gq family are not (Wu et al., 1993; Venkatakrishnan et al., 2000).
The formyl peptide receptors (FPRs) mediate chemotaxis activity. The three known FPRs are seven-transmembrane-domain GPCRs that are important in host defense and inflammation (Ye et al., 2009). FPR1 and FPR2 expression was first described in neutrophils and monocytes (Durstin et al., 1994). The expression profiles of FPR1 and FPR2 are similar. They were later observed in immature dendritic cells, microglial cells, spleen and bone marrow and some other tissues at lower levels (Migeotte et al., 2006). FPR3 transcripts are not found in neutrophils but can be detected in monocytes (Durstin et al., 1994). The tripeptide N-formyl-Met-Leu-Phe (fMLF) is a prototype of formylated chemoattractant peptides for neutrophils owing to its ability to bind and activate FPR1. At subnanomolar to nanomolar concentrations, this binding event translates into directional movement of neutrophils (Nanamori et al., 2004). Recently, a number of FPR ligands have been identified, such as the synthetic hexapeptides WKYMVM and WKYMVm, which activate FPR1 and FPR2. Annexin A1, a glucocorticoid-regulated protein, and its cleavage peptides Ac2-26 bind FPRs with affinities in the low micromolar range (Walther et al., 2000). These receptors also have other endogenous ligands, such as cathepsin G for FPR1, and lipoxin A4, SAA and LL37 for FPR2 (Migeotte et al., 2006).
Here, we employed a peripheral blood mononuclear cell (PBMC) chemoattractant platform to identify candidate chemoattractant cytokines and found that PBMCs and neutrophils exhibited chemotaxis toward family with sequence similarity 3 member D (FAM3D), a protein whose function is currently unknown. FAM3D belongs to a cytokine-like family that shares a four-helix-bundle structure (Zhu et al., 2002). The family consists of four members: FAM3A, FAM3B, FAM3C and FAM3D. FAM3D shows 53%, 28% and 50% amino acid sequence identity to FAM3A, FAM3B and FAM3C, respectively. FAM3B, also called pancreatic-derived factor, is an important regulator of glucose and lipid metabolism (Yang and Guan, 2013). FAM3D is constitutively expressed in the gastrointestinal tract and its function is linked to nutritional regulation (de Wit et al., 2012), but its chemotaxis function remains unknown. We found that human peripheral blood neutrophils and monocytes have strong chemotaxis activity to FAM3D. The chemoattractant receptors for FAM3D were determined to be FPR1 and FPR2 through functional screens for the classical and other chemoattractant receptors. Furthermore, we found that FAM3D expression increased significantly in dextran sulfate sodium (DSS)-induced colitis. It is known that chemokines and chemokine receptors play a key role in gastrointestinal homeostasis and inflammation (Griffith et al., 2014; Hill and Artis, 2010), and the endogenous FPR ligand, annexin A1, has been shown to regulate intestinal mucosal injury, inflammation and repair (Babbin et al., 2008). Therefore, we propose that FAM3D might function in gastrointestinal homeostasis and inflammation.
FAM3D is a classical secretory protein and strongly chemoattracts PMNs and PBMs
FAM3D belongs to the FAM3 family, which is a cytokine-like family. However, whether it is a classical secretory protein is unknown. We produced a secreted recombinant FAM3D protein through the PEI transfection system. The purified FAM3D protein exhibited a purity of at least 90% as assessed by an optical density assay (Fig. 1A). Furthermore, as shown in Fig. 1B, the secretion of FAM3D was inhibited by the Golgi blocker brefeldin A (BFA), suggesting that FAM3D is secreted through the classical ER–Golgi pathway. N-terminal sequencing showed that the mature protein sequence started at the 39th amino acid, with the first 38 amino acid residues being a signal peptide sequence (Fig. 1C, underlined). However, this was not consistent with the prediction made by SignalP-HMM software, which suggested a 25-amino-acid signal peptide sequence.
Because the supernatant of HEK293T cells transfected with pcDB-FAM3D induced the migration of PBMCs (Fig. S1), we wanted to know the leukocyte cell chemotaxis type. Thus, we separated polymorphonuclear neutrophils (PMNs), peripheral blood monocytes (PBMs) and peripheral blood lymphocytes (PBLs) from human peripheral blood. As shown in Fig. 1D, when we used the purified recombinant FAM3D protein to determine the chemoattractant effect, both PMNs and PBMs were strongly chemoattracted by FAM3D, but the chemoattraction to PBLs was not strong.
FAM3D is an agonist in FPR1- and FPR2-mediated chemotaxis
The major function of chemokines is to induce directional migration of cells expressing receptors for that chemokine. Therefore, a chemotaxis assay with HEK293 cells transiently transfected with different receptors was used to investigate the agonistic properties of FAM3D. FAM3D chemoattracted FPR1- and FPR2-expressing HEK293 cells (Fig. 2A,B), but HEK293 cells transiently expressing CXCR1 or CXCR2, both of which are highly expressed on the cell surface of neutrophils, were not chemoattracted by FAM3D (Fig. 2C,D). FPR3-overexpressing cells were not chemoattracted by FAM3D (Fig. S1). When we used cyclosporine H, a specific inhibitor of FPR1, the chemoattractant ability of both FAM3D and fMLF disappeared (Fig. 2A). As Fig. 2B shows, the chemoattractant ability of FAM3D and WKYMVm to FPR2 was inhibited by WRW4, an inhibitor of FPR2. To further evaluate the relationship between FAM3D and FPR, we prepared mouse monoclonal antibodies targeting FAM3D (Fig. S2) and found that the antibodies 5E12 and 6D7 effectively neutralized the ability of FAM3D to chemoattract FPR1-expressing HEK293 cells (Fig. 2E). Interestingly, only 6D7 effectively neutralized the ability of FAM3D to chemoattract FPR2-expressing cells (Fig. 2F), thus suggesting the binding sites of the two monoclonal antibodies to FAM3D are not exactly the same. Furthermore, we found that both FPR1 and FPR2 are expressed by mouse peritoneal neutrophils, and 6D7 could block the chemotaxis action of FAM3D completely, whereas 5E12 could not (Fig. S3). Because fMLF- or WKYMVm-induced responses require the physical association of FPR1 or FPR2 with a PTX-sensitive Gi/o protein in leukocytes (Bommakanti et al., 1992; Schreiber et al., 1993), we also investigated the chemoattractant activities of FAM3D to HEK293 cells expressing FPR1 and FPR2. Both were inhibited by PTX (Fig. 2G,H), which suggests that the pathway induced by FAM3D relies on the Gi/o protein.
FAM3D stimulates internalization of FPR1 and FPR2
GPCRs rapidly internalize upon ligand binding (von Zastrow, 2003). Based on this property, we examined FAM3D for its ability to bind to FPR1 or FPR2 and stimulate their internalization. We used recombinant 100 nM and 1000 nM FAM3D to stimulate HEK293 cells transiently expressing FPR1–EGFP for 1 h. Confocal microscopy showed that FAM3D and fMLF effectively stimulated the internalization of FPR1 (Fig. 3A). We next examined the surface expression of FPR1 on transfected HEK293 cells by flow cytometry. It reduced after 1 h of 100 nM or 1000 nM FAM3D stimulation (Fig. 3B). In comparison, fMLF, which is an agonist of FPR1, also induced the internalization of FPR1.
When we stimulated HEK293 cells overexpressing FPR2–EGFP with 10 nM or 100 nM FAM3D or WKYMVm for 1 h, receptor internalization was captured clearly by confocal microscopy (Fig. 4A). Next, the surface expression of FPR2 on the transfected HEK293 cells was examined by flow cytometry after stimulation with FAM3D or WKYMVm (Fig. 4B). We found that the expression of FPR2 was decreased by FAM3D or WKYMVm. The rate of internalization of FPR1 or FPR2 was calculated from three independent experiments, which showed that FAM3D strongly stimulated internalization of FPR1 and FPR2 (Figs 3C and 4C).
FAM3D can induce FPR1- or FPR2-mediated Ca2+ flux in HEK293 cells transiently transfected with FPR1 or FPR2
As FPR1 mediated fMLF-induced Ca2+ mobilization in a PTX-sensitive manner (Prossnitz et al., 1991), we examined whether FAM3D also affected intracellular Ca2+ levels in FPR1- or FPR2-transfected HEK293 cells. The cells were loaded with the Ca2+-sensitive fluorescent probe fluo-3 AM and stimulated either with FAM3D and fMLF or with FAM3D and WKYMVm. At 200 nM, FAM3D induced Ca2+ mobilization with a magnitude similar to that induced by 10 nM fMLF (Fig. 5A,B). Furthermore, 200 nM FAM3D and 10 nM fMLF cross-desensitized Ca2+ flux induced by each other (Fig. 5A,B). As shown in Fig. 5C,D, at 200 nM, FAM3D induced FPR2-mediated Ca2+ flux in FPR2-expressing HEK293 cells. The addition of both 200 nM FAM3D and 10 nM WKYMVm cross-desensitized Ca2+ flux induced by each other.
Radioactive binding assay confirms the receptors of FAM3D are FPR1 and FPR2
To further confirm that FPR1 and FPR2 are specific receptors for FAM3D, we performed a radioactive binding assay (RBA). After the recombinant mature FAM3D was labeled with 125I, we calculated a FAM3D–FPR1 saturation binding curve (Fig. 6A) by allowing 125I-FAM3D to react with varying quantities of membrane extract from HEK293 cells expressing FPR1–EGFP. We calculated a FAM3D–FPR2 saturation binding curve in the same way (Fig. 6B). A competitive binding assay was conducted with FAM3D, fMLF, cyclosporine H or CCL2 competing with 125I-labeled FAM3D to bind to FPR1. As shown in Fig. 6A, FAM3D, fMLF or cyclosporine H inhibited binding of 125I-FAM3D to FPR1. As a more potent agonist for FPR1, fMLF competed with the binding at a lower concentration. However, the uncorrelated control CCL2 did not compete with 125I-FAM3D. Similarly, in the FPR2 competitive binding assay, FAM3D, WKYMVm and WRW4 competed with 125I-FAM3D for binding to FPR2 to similar degrees, whereas the uncorrelated control RS102895, a specific CCR2B inhibitor, did not (Fig. 6B). As a ligand for FPR1, fMLF has a Kd of 0.04 nM (Fay et al., 1991); according to the ratio of the Ki for fMLF (Ki=0.1934 nM) to the Ki for FAM3D (Ki=34.46 nM) (calculated from the competitive binding analysis), the calculated dissociation constant Kd of FAM3D to FPR1 was 7.13 nM. As a ligand for FPR2, WKYMVm has a Kd of 1 pM (Migeotte et al., 2006); according to the ratio of the Ki for WKYMVm (Ki=21.53 nM) to the Ki for FAM3D (Ki=136.7 nM), the calculated dissociation constant Kd of FAM3D to FPR2 was 6.35 pM.
FAM3D can chemoattract neutrophils in peritoneal cellular recruitment
As FAM3D could chemoattract PMNs and PBMs in vitro (Fig. 1D; Movies 1–3), we wanted to know whether they could be recruited when FAM3D was injected into the peritoneal cavity of mice. At 6 h after intraperitoneal injection, compared to 14.0% CD11b+ Ly6G+ (CD11b is also known as ITGAM) neutrophils in control (Fig. 7A), 46.6% CD11b+ Ly6G+ neutrophils were collected in the FAM3D group (Fig. 7B). Furthermore, 86.8% of the neutrophils chemoattracted by FAM3D were Fpr2+, and 19.7% were Fpr1+ (Fig. 7C), and 82.6% of CD11b− Ly6G− cells were Fpr1− Fpr2− (Fig. 7D). Then, we calculated the relative numbers of various types of leukocytes, including neutrophils, T lymphocytes (CD3), B cells (B220), dendritic cells (CD11c) and macrophages (F4/80) upon intraperitoneal injection of FAM3D (Fig. 7E). Except for neutrophils, the number of other leukocytes had no obvious change.
ERK1/2 and p38 MAPK signaling is crucial for FAM3D-induced neutrophil activation
Activation of neutrophils through formyl peptide receptors depends on MAPK signaling (Hauser et al., 2010). However, whether the activation of the MAPKs is required for FAM3D-induced neutrophil activation and function has not been determined. After peritoneal neutrophils were isolated from the peritoneal cavity after thioglycollate injection, we pretreated freshly isolated neutrophils with cyclosporine H and/or WRW4 before FAM3D stimulation. We found that FAM3D stimulated an increase in the amount of phosphorylated ERK1 and ERK2 (ERK1/2, also known as MAPK3 and MAPK1, respectively) and the amount of phosphorylated p38 MAPK family proteins after 2 or 5 min (Fig. 7F). When neutrophils were pretreated with cyclosporine H, a selective antagonist of formyl peptide receptor 1, the upregulation of phosphorylated ERK1/2 did not change, but that of phosphorylated p38 MAPK family proteins was inhibited. When the cells were pretreated with WRW4, a selective antagonist of FPR2, FAM3D could no longer upregulate either phosphorylated ERK1/2 or phosphorylated p38 MAPK family proteins (Fig. 7F). This suggests the activation of neutrophils by FAM3D occurs via p38 MAPK family proteins and ERK1/2, and that the activation of ERK is dependent on FPR2, but the activation of p38 MAPK family proteins is both FPR1 and FPR2 dependent.
FAM3D is upregulated in the colon of mice in a DSS model
FAM3D expression was evaluated in multiple tissue cDNA Panels from Clontech Laboratories by real-time PCR. In general, FAM3D had a high level of expression in tonsil, a moderate level of expression in lung and pancreas, and a low level of expression in other human tissues (Fig. 8A). It has been reported previously that FAM3D is expressed in gastrointestinal tract tissues and as a gut-secreted protein displaying nutritional-status-dependent regulation (de Wit et al., 2012). To explore the role of FAM3D under physiological and pathological status, we detected FAM3D in mouse DSS-induced colitis by immunohistochemistry and found that its expression increased substantially (Fig. 8B). Real-time PCR confirmed this result (Fig. 8C). As the DSS-induced colitis model is widely used because of its simplicity and many similarities with human ulcerative colitis (Chassaing et al., 2014), FAM3D might be related to gastrointestinal inflammation processes.
Since the first chemokine, CXCL8, was described in 1987, 48 chemokines have been identified in humans, which makes this group the largest subfamily of cytokines (Zlotnik and Yoshie, 2012). They are directly involved in the migration and activation of leukocytes, especially phagocytes and lymphocytes, playing an important role in the inflammatory response. Several chemokine-like function (CLF) chemokines cannot be classified into known chemokine subfamilies. For example, MIF (Bernhagen et al., 2007), β-defensins (Yang et al., 1999) and a tyrosyl-tRNA-synthetase fragment (Wakasugi and Schimmel, 1999), among others, are also involved in leukocyte chemotaxis and the migration of normal cells and tumor cells.
In our laboratory, using a PBMC chemoattractant platform to screen for unknown function cytokines with chemokine function, we found that FAM3D exhibits chemoattractant ability. PC3-secreted microprotein (PSMP, also known as MSMP) exerts its chemoattractant activity through CCR2, which was also identified by our platform (Pei et al., 2014). Here, we found that FAM3D was classically secreted to yield a mature amino acid form. Purified recombinant FAM3D protein strongly induced the chemotactic migration of PMNs and PBMs, which might be the target cells of FAM3D. Furthermore, FAM3D chemoattracted numerous CD11b+ Ly6G+ neutrophils within 6 h in vivo, which is consistent with the strong induction of chemotactic migration of PMNs mediated by FAM3D in vitro.
Because the amino acid structure of mature FAM3D does not have the typical structural domain of classical chemokines, FAM3D seems to be a new CLF. Chemokines always exert their function by binding the GPCR receptors on the surface of immune cells. We decided to search for potential receptors of FAM3D through a Boyden chamber chemotaxis assay and found that FAM3D induced the chemotaxis of HEK293 cells expressing FPR1 or FPR2, but that CXCR1- or CXCR2-expressing HEK293 cells were not chemoattracted by FAM3D. In addition, FAM3D could not chemoattract cells expressing many other classical chemokine receptors. The results of chemotaxis, receptor internalization, Ca2+ flux and RBA assays led us to conclude that FAM3D is a new agonist of FPR1 and FPR2. Additionally, the chemotaxis phenotype was inhibited by an inhibitor of the relevant receptor. Furthermore, when we used monoclonal antibodies against FAM3D (5E12 and 6D7), the chemoattractant activity of FPR1 was neutralized by 5E12 and 6D7. By contrast, the chemoattractant activity of FPR2 was neutralized by 6D7 completely, but not by 5E12. These data show that FAM3D has a ligand–receptor relationship with FPR1 and FPR2. Finally, the chemoattractant activity of FAM3D to HEK293 expressing FPR1 or FPR2 was inhibited by PTX, thus suggesting the pathway induced by FAM3D is dependent on Gi/o protein. As FPR1 and FPR2 are similarly highly expressed on the cell surface of neutrophils and monocytes, but are expressed to a lesser degree in the lymphocytes (Migeotte et al., 2006), they might not be the only two receptors of FAM3D.
FPR1 and FPR2 are mainly expressed in neutrophils and monocytes, and they have been known to transduce proinflammatory (Karlsson et al., 2009; Partida-Sanchez et al., 2004) as well as anti-inflammatory signals (Herrera et al., 2015; Wang et al., 2015). During a response to infection, accumulation and activation of neutrophil granulocytes at inflammatory sites are of profound importance (Bidula et al., 2015). FPR-related signaling pathways, including MAPK (Hazeldine et al., 2015), PI3K–AKT (Wang et al., 2015), and JAK–STAT–SOCS (Pupjalis et al., 2011) pathways in neutrophils or macrophages, participate in the proliferation, metastasis and/or survival of these leukocytes. Our results indicate that FAM3D activates ERK1/2 and p38 MAPK proteins in neutrophils through FPR1 or FPR2.
Formylated peptides produced by bacteria are high-affinity exogenous ligands of FPR1 and FPR2. They play crucial roles in infectious diseases (Molloy et al., 2013). Other high-affinity ligands of FPR2 include WKYMVm and WKYMVM from random peptide libraries. Like FAM3D, annexin A1 and its N-terminal peptide Ac 2-26 are also endogenous ligands of FPR1 and FPR2 (Ernst et al., 2004). The endogenous peptide Ac2-26 is cleaved from annexin A1, but its Kd for FPR1 or FPR2 is in the low micromolar range (Migeotte et al., 2006). In acute and chronic inflammation, annexin A1 is activated during the process of neutrophil extravasation (Oliani et al., 2001). It also mediates the rapid anti-inflammatory effects of neutrophil-derived microparticles (Dalli et al., 2008). Annexin A1 is expressed in many organs of the body, and highly in the intestinal tract. In the DSS-induced colitis model, annexin A1−/− mice are more susceptible to developing enteritis and recover slowly than wild-type mice when the DSS is withdrawn (Babbin et al., 2008), which suggests that annexin A1 regulates intestinal mucosal injury, inflammation and repair. WKYMVm, a FPR2 synthetic agonist from a random peptide library, reverses decreases in body weight, bleeding score and stool score in DSS-treated mice; these functions are inhibited by WRW4 (Kim et al., 2013). Chemokines and chemokine receptors play a key role in gastrointestinal homeostasis and inflammation (Griffith et al., 2014; Hill and Artis, 2010). As an endogenous ligand of FPR1 and FPR2, annexin A1 regulates intestinal mucosal injury, inflammation and repair (Babbin et al., 2008; Fay et al., 1991). Fpr2−/− mice also show diminished PMN infiltration into the colonic mucosa in acute DSS-induced colitis, delayed mucosal restoration after injury and increased azoxymethane-induced tumorigenesis, suggesting that FPR2 contributes to colonic epithelial homeostasis, inflammation and tumorigenesis (Chen et al., 2013). FAM3D is highly expressed in the gastrointestinal tract (de Wit et al., 2012) and our data demonstrated that FAM3D was upregulated in colon tissue following DSS treatment. Therefore, we propose that FAM3D might play a role in gastrointestinal-related inflammation processes.
In conclusion, our work identifies FAM3D as a new chemotaxis agonist for the G protein-coupled formyl peptide receptors FPR1 and FPR2. Moreover, FAM3D can activate ERK1/2 and p38 MAPK signaling in neutrophils through FPR1 or FPR2. As FAM3D is constitutively expressed in the gastrointestinal tract and upregulated in DSS-induced colitis, FAM3D might play an important role in intestinal homeostasis and gastrointestinal inflammation.
MATERIALS AND METHODS
All mice used in this study were between 6–8 weeks old and of a C57BL/6 background and were kept at the Peking University Health Science Center under specific pathogen-free conditions. The animal experimental procedures were approved by the ethics committee of the Peking University Health Science Center.
Reagents and chemicals
RPMI 1640 medium and fetal bovine serum (FBS) were purchased from Life Technologies. Dulbecco's Modified Eagle's medium (DMEM), Brefeldin A (BFA), bovine serum albumin (BSA), thioglycollate, polyethyleneimine (PEI) and mouse IgG were purchased from Sigma-Aldrich. Mouse monoclonal antibody against His (clone OGHis) for western blotting were obtained from Medical & Biological Laboratories Co. (Japan) and used at 1:1000. Pertussis toxin (PTX) was purchased from Alexis Biochemical Corporation. 125INa was obtained from DuPont. CXCL8, CXCL12, CCL2 and CCL23 were purchased from PeproTech Inc. Fluo-3 AM was obtained from Invitrogen. Goat anti-FAM3D (no. AF2869, 1:100), mouse anti-FPR1 (no. MAB3744, 1:100), anti-FPR2 (no. MAB3479, 1:100), phycoerythrin-conjugated anti-FPR1 (no. FAB3744P, 1:100), allophycocyanin (APC)-conjugated anti-FPR2 (no. FAB3479A, 1:100) and fMLF, WKYMVm, WRW4, RS102895 were obtained from R&D Systems. Cyclosporine H was produced at Fujian Institute of Microbiology (Fuzhou, China) as described previously (Yan et al., 2006). FITC-conjugated anti-mouse CD45 (clone 30-F11, 1:500), FITC-conjugated anti-mouse CD11b (clone M1/70, 1:500), phycoerythrin-conjugated anti-mouse B220 (clone RA3-6B2, 1:500), APC-conjugated anti-mouse F4/80 (clone BM8, 1:500), APC-conjugated anti-mouse CD11c (clone N418, 1:500) were purchased from Biolegend, phycoerythrin-conjugated anti-mouse IgG (clone M1-14D12, 1:500), Percp/cy5.5-conjugated anti-mouse CD3 (clone 145-2C11, 1:500) and Percp/cy5.5-conjugated anti-mouse Ly6G (clone RB6-8C5, 1:500) were purchased from eBioscience. The protease inhibitor cocktail and the PhosSTOP cocktail were obtained from Roche (Rotkreuz, Switzerland). Rabbit monoclonal antibodies against p38 (no. 8690, 1:1000) and phospho-p38 (No. 4511, 1:1000) and ERK1/2 (clone 4695) and mouse monoclonal antibodies against phospho-ERK1/2 (clone 9106, 1:1000) for the western blotting were obtained from Cell Signaling Technology.
The pcDNA3.1-FPR2 and pCEP4-CXCR1 expression plasmids were kindly provided by Professor Philip M. Murphy from the Laboratory of Molecular Immunology, National Institutes of Health, Bethesda, MD. The pcDNA3.1-FPR1, pcDNA3.1-CXCR2, pEGFP-N1-FPR1, pEGFP-N1-FPR2 and pcDNA3.1-FAM3D-myc-his plasmids were constructed in our laboratory.
HEK293 and HEK293T cells were purchased from ATCC (Manassas, VA) and were cultured in complete DMEM with 10% FBS (or inactivated FBS for the latter).
Production of recombinant human FAM3D protein
The FAM3D DNA sequence was inserted into the pcDNA3.1-myc-his plasmid. The pcDNA3.1-FAM3D-myc-his plasmid was transfected with polyethyleneimine (PEI) into the HEK293T cell line according to the manufacturer's recommended protocol. At 16 h after transfection, HEK293T culture medium was replaced with Hektor S medium (Cell Culture Technologies, Switzerland) with 2% glutamate (Sigma-Aldrich). The culture medium of the transfected HEK293T cells was collected at 120 h after the replacement of medium. Recombinant human FAM3D was purified using the Ni-Sepharose High Performance system according to manufacturer's instructions (GE Healthcare). Finally, FAM3D protein was eluted by 500 mM imidazole in 10 ml buffer (pH 7.4), and high-concentration imidazole was replaced with phosphate-buffered saline (PBS, pH 7.2) by ultrafiltration device (3000 m.w., Millipore, Billerica, MA). All glassware used to purify FAM3D was pretreated at 180°C for 4 h to eliminate endotoxin. The recombinant FAM3D was dissolved in PBS (pH 7.2) at 400 μg/ml and stored at −80°C.
The purified recombinant FAM3D protein was separated by SDS-PAGE and transferred onto a hydrophobic polyvinylidene fluoride (PVDF) membrane (Whatman). Then, the membrane was stained with Coomassie Blue, and the N-terminal sequencing was performed by Shanghai GeneCore BioTechnologies Co., Ltd.
Separation of PBMCs and PMNs from human peripheral blood cells
Human peripheral blood was obtained from the Beijing Red Cross Blood Center. Peripheral leukocytes were separated by Polymorphprep (Axis-Shield, Norway) at a 1.077 density following the manufacturer's instructions. PBMCs were located at the interface between RPMI 1640 and Polymorphprep, and polymorphonuclear neutrophils (PMNs) were located at the interface between Polymorphprep and red blood cells. After washing PBMCs and PMNs twice with RPMI 1640, the PBMCs were cultured in RPMI 1640 with 10% inactivated FBS. Then, the PMNs were used in the chemotaxis assay. After culture overnight, the PBMCs were separated into peripheral blood monocytes (PBMs) and peripheral blood lymphocytes (PBLs). PBMs grew attached to the dish, PBLs grew in suspension. The PBMs and PBLs were washed separately, at which point they were tested for their chemoattraction to FAM3D in Boyden chambers.
Preparation of mouse anti-FAM3D monoclonal antibodies
The FAM3D eukaryotic protein and prokaryotic protein were used to immunize mice. After screening hundreds of monoclonal antibodies, we choose several of the best specific antibodies for the experiments and purified them by means of protein-G.
Receptor-transfected HEK293 cells
HEK293 cells (ATCC) were grown in RPMI 1640 supplemented with 10% FBS, penicillin (100 U/ml) and streptomycin (100 μg/ml). 4×106 HEK293 cells in 400 μl were transiently transfected by electroporation with 20 μg of the expression plasmid at 120 V for 20 ms using an electric pulse generator (Electro Square Porator ECM 830, BTX, SanDiego, CA) and chemotaxis assays, receptor internalization, Ca2+ mobilization and radioligand binding assays were performed 48 h later.
The chemotaxis assay was performed using a 48-well microchemotaxis chamber (Neutroprobe; Cabin John, MD) as described previously (Sarafi et al., 1997). In brief, chemoattractants were diluted in RPMI 1640 medium supplemented with 0.5% BSA and placed in the lower wells (27 μl/well). PMN, PBM, PBL or HEK293 cells transfected with plasmid were resuspended in the same medium at 1×106 cells/ml and added to the upper wells (55 μl/well), which were separated from the lower wells by a polyvinylpyrrolidone-free polycarbonate filter with pores. After incubation at 37°C in 5% CO2 for several hours, cells that had migrated into the lower chamber were counted by a Veritas Microplate Luminometer (Turner Systems) or the membranes were stained by the three-step staining buffer. The chemotaxis index was calculated from the number of cells that migrated compared to the control. Significant chemotaxis was defined as chemotaxis index>2.
Receptor internalization assay
We used confocal microscopy to screen the receptor internalization directly. An FPR1–EGFP or FPR2–EGFP construct was prepared by ligation of FPR1 or FPR2 cDNA to the N-terminus of EGFP (BD Biosciences Clontech, Palo Alto, CA). A HEK293 cell line expressing FPR1–EGFP or FPR2–EGFP was generated by transient transfection as described above. For the internalization assay, the HEK293 cells expressing FPR1–EGFP or FPR2–EGFP cells were grown on glass coverslips for 16 h in RPMI 1640 supplemented with 10% FBS. The transfected cells were stimulated with FAM3D, fMLF, WKYMVm or BSA for 1 h in supernatant. Internalization was terminated by adding fixation buffer (3% paraformaldehyde in PBS) followed by incubation at room temperature for 15 min. The cells were then washed twice with PBS, and the diaminobenzidine detection system (DAKO code k5007) was used to visualize expression. The subcellular localization of EGFP was observed under a confocal microscope (Leica TCS SP8).
We also used flow cytometry to detect the receptor internalization. After electro-transfecting pcDB-FPR1 or pcDB-FPR2 into HEK293 cells as described above, HEK293 cells expressing FPR1 or FPR2 cells were stimulated with fMLF or WKYMVm, FAM3D, or BSA in suspension for 1 h. The cells were then washed with PBS twice and incubated in PBS with 20% FBS and 1 μg/ml human IgG for blocking Fc on the cell surface. Then, cells were washed with cold PBS and incubated with anti-FPR1, anti-FPR2 or mouse IgG primary antibody at 4°C for 1 h. After being washed twice in PBS, the cells were incubated with goat phycoerythrin-conjugated anti-mouse-IgG as the detection antibody for 30 min, then washed twice in PBS and resuspended in cold PBS. Finally, 10,000 cells per experiment were analyzed by flow cytometry using a FACSCalibur, the CellQuest software (BD Biosciences), and the FlowJo 7.6.1 software.
Ca2+ flux assay
After electro-transfection with pcDB-FPR1 or pcDB-FPR2, HEK293 cells were cultured in 35-mm confocal dishes for 48 h. Then, the cells were incubated with 10 μM fluo-3AM in RPMI 1640 at 37°C for 1 h in the dark and examined for Ca2+ mobilization in response to the compounds using a Leica SP8 confocal laser-scanning microscope. HEK293 cells were loaded with fluo-3AM, washed twice with PBS, and stimulated by 200 nM FAM3D, 10 nM fMLF or 10 nM WKYMVm. Ca2+ flux changes were recorded using a Leica SP8 confocal laser-scanning microscope; images were acquired every 5 s, and the first 40 s were used as the basal Ca2+ flux line before any stimulation was added. Ca2+ flux was recorded for 3 min before the second stimulation. Ca2+ flux intensity was assessed by Leica System Analysis software.
Radioligand binding assay
First, we established an HEK293 cell line stably expressing FPR1–EGFP or FPR2–EGFP. FAM3D was labeled with 125I. In the saturation experiment, 125I-FAM3D was incubated with varying quantities of the FPR1 or FPR2 cell membrane extract in binding buffer. In the competition experiment, 125I-FAM3D was incubated with varying quantities of unlabeled FAM3D, fMLF, WKYMVm, CCL2 or RS102895. The radioactivity assay was evaluated as described previously (Pei et al., 2014).
Peritoneal cellular recruitment
Cellular recruitment was induced by intraperitoneal injection of female C57BL/6 mice (6–8 weeks old) with 10 μg of FAM3D protein diluted in 0.1 ml of sterile, lipopolysaccharide (LPS)-free PBS. Basal peritoneal cells were counted by the administration of 0.1 ml of sterile, LPS-free PBS. After 6 h, mice were killed and peritoneal lavage was performed with three washes of 5 ml of sterile, ice-cold PBS. Samples were centrifuged at 600 g for 10 min. The cells were resuspended in a final volume of 0.5 ml, and the total leukocytes were incubated with specific antibodies to identify the cell types as follows: FITC-conjugated anti-mouse CD11b and Percp/cy5.5-conjugated anti-mouse Ly6G for neutrophils, FITC-conjugated anti-mouse CD45 for leukocytes, APC-conjugated anti-mouse F4/80 for macrophages, APC-conjugated anti-mouse CD11c for dendritic cells, phycoerythrin-conjugated anti-mouse B220 for B lymphocytes, Percp/cy5.5-conjugated anti-mouse CD3 for T lymphocytes. All antibodies against CD45, CD11b, Ly6G, FPR1, FPR2, F4/80, CD11c, CD3 and B220 were used in accordance with the manufacturers’ instructions. Finally, 10,000 leukocytes per experiment were analyzed by flow cytometry using a FACSCalibur, CellQuest software (BD Biosciences) and FlowJo 7.6.1 software.
Preparation of peritoneal neutrophils and stimulated it by FAM3D
For the preparation of peritoneal neutrophils, cells were isolated from the peritoneal cavity 4 h after 1 ml 4% thioglycollate injection and cultured for 18 h. Then, nonadherent cells were harvested and stimulated as described previously (Takeda et al., 1999). The neutrophils were stimulated by 200 ng/ml FAM3D, or before stimulating, cells were treated with 5 μM FPR1 inhibitor cyclosporine H or the 5 μM FPR2 inhibitor WRW4 for 1 h. After being stimulated 2 min or 5 min, the cell lysates were obtained for western blot.
Cell lysates were extracted with ice-cold RIPA lysis buffer for 30 min and cleared by centrifugation at 6000 g for 5 min at 4°C. Total protein concentration was measured using a bicinchoninic acid protein assay kit (Pierce). Proteins were resolved by SDS-PAGE and transferred onto nitrocellulose membranes (HybondTMECLTM, Amersham Biosciences). After blocking the nitrocellulose membranes in ODYSSEY blocking buffer for 1 h, membranes were incubated with primary antibodies at 4°C overnight and then with ODYSSEY 700/800-labeled anti-IgG secondary antibody at room temperature for 1 h in the dark. Finally, the fluorescence intensity on the membranes was detected with the LI-COR infrared imaging system and was analyzed using ODYSSEY software.
The expression of FAM3D mRNA was examined using real-time PCR with the UPL probes and GAPDH as the internal reference. The following primers were used: FAM3D-F, 5′-GTAAAAGCCCCTTTGAGCAGT-3′; FAM3D-R, 5′-GGCCATCCCTCGTATTTGT-3′; Fpr1-F, 5′-TGTCCAGAGCTGTTGGAAAGT-3′; Fpr1-R, 5′-TTCATGAGGTTCACTGCAGACT-3′; Fpr2-F, 5′-CCACAGGAACCGAAGAGTGT-3′; Fpr2-R, 5′-CCACCACTTCTGATCCATTCA-3′; GAPDH-F, 5′-TCCACTGGCGTCTTCACC-3′; GAPDH-R, 5′-GGCAGAGATGATGACCCTTTT-3′.
Real-time PCR was performed in the ABI Prism 7000 Sequence Detection System (Applied Biosystems, Foster City, CA) using the human Universal Probe Library system (Roche) and Taqman Gene Expression Master Mix (Applied Biosystems). The samples were run in triplicate with the following cycling conditions: 10 min denaturing at 95°C, followed by 40 cycles of 15 s at 95°C and 1 min at 60°C.
Tissue paraffin sections were pretreated in 70°C for at least 3 h. After soaking in dimethylbenzene for 15 min, the sections were hydrated in 100%, 95%, 90%, 80%, 75% and 70% ethanol. Antigen retrieval was performed using a high-pressure method in citrate buffer (pH 6.0). Peroxidase was removed by 3% H2O2 for 10 min. Samples were blocked with rabbit serum in PBS for 40 min at room temperature, and 1 μg/ml goat anti-FAM3D antibody or anti-IgG antibody was applied to the section for 1 h at 37°C. The tissue section was washed three times in PBS (pH 7.2), and then anti-goat horseradish-peroxidase-conjugated second antibody was applied for 30 min. All antibodies were diluted in PBS. After washing three times in PBS, the diaminobenzidine detection system (DAKO code k5007) was used to visualize expression.
DSS-induced experimental colitis in mice
Ulcerative colitis was induced by feeding C57BL/6 mice (6 weeks old) 3% (w/v) DSS (molecular mass 36,000–50,000 Da; MP Biomedicals, Solon, OH) in drinking water. The mice were given free access to water containing DSS for 7 days. Control mice received water without DSS. The mice were weighed and monitored for the appearance of diarrhea and blood in the stool throughout the experimental period.
Statistical analysis was performed using the two-tailed Student's t-test (unpaired) for determining differences between two samples. Data were expressed as mean±s.e.m. Significant differences between groups are represented by *P<0.05, **P<0.01 and ***P<0.001. Comparative analysis was analyzed in Prism 5.0 (GraphPad Software).
We thank Philip M. Murphy (Laboratory of Molecular Immunology, National Institute of Allergy and Infectious Diseases, National Institutes of Health, Bethesda, MD) for kindly providing FPR2 and CXCR1 expression plasmids.
X.J.P. and E.Q.X. conceived or designed the experiments, performed experiments and wrote the manuscript. W.W.L. performed the experiments and analyzed the data. X.L.P., D.X.C. and D.F.Z. analyzed the data. Y.Z., C.Z., S.P.S. and J.M. interpreted the data. P.Z.W. screened candidate chemokine genes in the biological information analysis. Y.Z. established the screening platform for the chemokines. X.N.M. finished the work of real-time PCR and analyzed this data. Y.M.Z. gave important advice for culturing cells. D.L.M. provided important advice for the work. Y.W. conceived or designed the experiments and wrote the manuscript.
This work was supported by National Key Basic Research Program of China [grant number 2012CB518002 partly to Y.W.], National Natural Science Foundation of China [grant numbers 31270915 and 31470842 to Y.W.]; Beijing Municipal Natural Science Foundation [grant number 7152082 to Y.W.], and Specialized Research Fund for the Doctoral Program of Higher Education of China [grant number 20120001110001 to Y.W.].
The authors declare no competing or financial interests.