C5a- and ASP-mediated C5L2 activation, endocytosis and recycling are lost in S323I-C5L2 mutation

Molecular Immunology 46 (2009) 3086–3098

Contents lists available at ScienceDirect

Molecular Immunology journal homepage: www.elsevier.com/locate/molimm

C5a- and ASP-mediated C5L2 activation, endocytosis and recycling are lost in S323I-C5L2 mutation Wei Cui a,d , May Simaan c , Stephane Laporte c,d , Robert Lodge b , Katherine Cianflone a,d,∗ a

CRIUCPQ, Université Laval, Québec, Canada CRCHUL, Université Laval, Québec, Canada Department of Pharmacology, McGill University, Montreal, Canada d Department of Experimental Medicine, McGill University, Montreal, Canada b c

a r t i c l e

i n f o

Article history: Received 22 April 2009 Received in revised form 26 May 2009 Accepted 4 June 2009 Available online 16 July 2009 Keywords: C3adesArg Rab proteins ␤-Arrestin-2 G-protein-coupled receptor C5L2

a b s t r a c t C5L2, a G-protein-coupled receptor (GPCR), has been identified as an ASP (C3adesArg) and C5a receptor. Controversy exists regarding both ligand binding and functionality. ASP activation of C5L2 is proposed to regulate fat storage. C5L2 is also proposed as a decoy receptor for C5a, an inflammatory mediator, based on absence of Ca2+ or chemotaxis changes. Aims: (i) to evaluate C5L2 receptor activation and recycling using recombinant ASP (rASP) and rC5a and (ii) assess receptor trafficking of S323I-C5L2 mutation previously identified in a family and demonstrated to have altered functionality. Results: stably transfected C5L2-HEK cells were sorted using fluorescent-ASP (Fluos-ASP) binding. Following 2-h serum-free pretreatment, C5L2 was typically localized to the cell-surface. ␤-Arrestin-2-GFP transiently transfected C5L2-HEK cells demonstrated rASP and rC5a-dependent ␤-arrestin-2-GFP translocation, which showed time-dependent intracellular colocalization with C5L2. Without ligand or C5L2 transfection, no translocation was identified at any time point. Ligand-dependent (rASP and rC5a) C5L2 endocytosis was time-dependent with a 1-h nadir, and was clathrin- and cholesterol-dependent. Transiently transfected Rab-GFP proteins (Rabs 5, 7 and 11) demonstrated time-dependent colocalization of Rab5, Rab7, and Rab11 with C5L2. In contrast to C5L2, a large proportion of stably transfected S323I-C5L2 did not localize to the cell-surface. While S323I-C5L2 was competent for Fluos-ASP and 125 I-ASP binding, although at a reduced level, there was no ligand-mediated receptor phosphorylation. Further, there was no ligand-mediated activation of ␤-arrestin-2-GFP translocation, and no downstream functional activation of glucose transport or triglyceride synthesis. Conclusion: C5L2 is a functional metabolic receptor, and serine 323 is important for ASP induced functionality. © 2009 Elsevier Ltd. All rights reserved.

1. Introduction C5L2, first identified as an orphan receptor by Ohno et al. (2000) and Lee et al. (2001), belongs to the seven transmembrane Gprotein-coupled receptor (GPCR) superfamily and consists of 337 amino acids (37 kDa) (Ohno et al., 2000). C5L2 shares 38%, 33%, 29%, and 29% amino acid identity with C5a, C3a, fMLP, and ChemR23

Abbreviations: ASP, acylation stimulating protein; FITC, fluorescein isothiocyanate; GFP, green fluorescent protein; GPCR, G-protein-coupled receptor; HA, hemaglutinin; HSF, human skin fibroblast; KO, knockout; LPS, lipopolysaccharide; OVA, ovalbumin; PFA, paraformaldehyde; TG, triglyceride; TGN, trans-golgi network; TLC, thin layer chromatographic; TRITC, tetramethylrhodaminoisothiocyanate. ∗ Corresponding author at: Centre de Recherche Institut Universitaire de Cardiologie & Pneumologie de Québec, Université Laval, Y2186, 2725 Chemin Ste-Foy, Québec, QC, Canada G1V 4G5. Tel.: +1 418 656 8711x3731; fax: +1 418 656 4602. E-mail address: katherine.cianfl[email protected] (K. Cianflone). 0161-5890/$ – see front matter © 2009 Elsevier Ltd. All rights reserved. doi:10.1016/j.molimm.2009.06.007

receptors, respectively (Ohno et al., 2000). C5L2 is expressed in many tissues, including all adipose tissues, such as pectoral, perirenal, gonadal and inguinal fat depots, and brown adipose tissue, in addition to brain, kidney, liver, spleen, intestine and myeloid cells (Kalant et al., 2005). The function of C5L2 remains controversial with questions regarding: (i) ligand binding, (ii) activation and functional response and (iii) impact in C5L2 knockout animal models. Both C5a and C5adesArg have been shown to bind C5L2, but with differing affinities (Cain and Monk, 2002). A functional receptor for C5a has been previously identified as CD88 (C5aR), which is a 39 kDa protein (350 amino acids) member of the anaphylatoxin receptor family which belongs to the rhodopsin-like GPCR superfamily (Gerard and Gerard, 1991). C5aR is expressed in granulocytes, macrophages, mast cells and dendritic cells (Chenoweth and Goodman, 1983). In addition, it has been shown that most organs express C5aR such as smooth muscle, liver, lung, kidney,

W. Cui et al. / Molecular Immunology 46 (2009) 3086–3098

spleen, small intestine, and heart (Fayyazi et al., 2000). C5a binding to C5aR induces inflammatory effects such as chemotaxis and oxidative burst (Burg et al., 1996), and ␤-hexosaminidase release (Cain and Monk, 2002). C5adesArg binds with lower affinity, resulting in decreased or absence of response (Perez, 1984). It has been well documented that stimulation of C5aR with C5a induces rapid phosphorylation (<2 min) and subsequent internalization (Giannini and Boulay, 1995). By 5 min post-stimulation, 60% of C5aR is internalized and after 10 min the internalized receptors within endosomes localize to the perinuclear region (Naik et al., 1997). Using analogous methodology and time frame, several studies have consistently shown that C5a binding to C5L2 did not result in stimulation of ␤-hexosaminidase release or increased intracellular calcium concentration (Cain and Monk, 2002; Ohno et al., 2000). These two criteria have typically been used to identify decoy chemokine receptors (Borroni et al., 2008). Accordingly, it has been

3087

suggested that C5L2 does not interact with G-proteins and acts as a decoy receptor for C5a. However there are many downstream effectors of G-proteins which are Ca2+ independent (Brambilla et al., 2002; Kehrl, 1998), and this alone cannot be used to exclude a functional role for C5L2. In fact, a body of evidence does suggest C5L2 functionality. C5a activation of C5L2 does result in phosphorylation of C5L2 (Okinaga et al., 2003), enhances the IgE effect on ␤-hexosaminidase release (Cain and Monk, 2002) and induces C5L2 internalization (Kalant et al., 2005), albeit on a slower time scale than with C5aR. Furthermore, macrophages and neutrophils from C5L2 knockout mice had decreased or absent C5a- and/or C3a-mediated Akt phosphorylation and ERK1/2 phosphorylation (in spite of expression of C5aR and C3aR) (Chen et al., 2007). Acylation stimulating protein (ASP, also known as C3adesArg) is an anabolic hormone which regulates fat storage in tissues

Fig. 1. rASP and rC5a induce ␤-arrestin-2-mediated C5L2 internalization. C5L2-HEK cells were seeded onto glass coverslips and transfected with ␤-arrestin-2-GFP. At 48 h post-transfection, cells were preincubated in serum-free media for 2 h, followed by ligand stimulation with 600 nM rASP (A) or 50 nM rC5a (B) for 0, 5, 15, 60, and 90 min. After stimulation, cells were: (i) fixed in 4% PFA–PBS for 8 min, (ii) permealized with 0.2% Triton in 4% BSA–PBS for 15 min, followed by rabbit anti-HA antibody incubation for 1.5 h at room temperature. After washing, cells were incubated with goat anti-rabbit antibody conjugated with TRITC at room temperature for 1 h. Cells were washed with PBS before the coverslips were mounted on microscope slides. The samples were dried overnight in the dark and visualized using confocal microscopy.

3088

W. Cui et al. / Molecular Immunology 46 (2009) 3086–3098

Fig. 1. (Continued ).

by stimulating triglyceride (TG) synthesis and glucose transport (Kalant et al., 2005). ASP (C3adesArg) and C3a have also been shown to bind C5L2 (Cain and Monk, 2002; Kalant et al., 2003, 2005). By contrast, C3a, but not C3adesArg (ASP), binds to the C3a receptor (C3aR) (Kohl, 2001), and neither C3a nor ASP binds to C5aR (Kalant et al., 2005). In C5L2-HEK stably transfected cells sorted through fluorescent-ASP binding, ASP stimulates TG synthesis and glucose transport, while there is no effect in HEK 293 cells (Kalant et al., 2005). It has also been demonstrated that ASP-mediated C5L2 activation induced C5L2 phosphorylation, and increased ␤-arrestin-2-GFP translocation to the cell-surface membrane followed by enhanced formation of ␤-arrestin-2-GFP associated intracellular endosomes (Kalant et al., 2005). In the absence of ligand or C5L2, there was no change in ␤-arrestin-2GFP distribution (Kalant et al., 2005). In addition, loss-of-function using antisense or si-RNA in C5L2 endogenously expressed in 3T3L1 preadipocytes and human skin fibroblasts (HSF) cells resulted in a decrease in TG synthesis, coordinate with a reduction of C5L2 expression (Kalant et al., 2005). Furthermore, in 3T3-L1 cells, ASP induces ERK1/2 phosphorylation and Akt phosphorylation (Maslowska et al., 2006), consistent with the results in C5L2 KO mice described above (Chen et al., 2007), as well as phospholipase A2 and protein kinase C activation (Baldo et al., 1995; Maslowska et al., 2006). Blocking these signalling pathways blocks the ASP induced stimulation of TG synthesis (Maslowska et al., 2006), and in C5L2 KO mice, ASP no longer stimulates TG synthesis in adipocytes (Paglialunga et al., 2007). Thus the only functional receptor identified to date for ASP is C5L2.

As a lipogenic adipokine, ASP plays a potent role in post-prandial clearance in vivo and is linked to hyperlipidemic disorders, characterized by increased plasma triglycerides or apolipoprotein B containing lipoproteins (Cianflone et al., 2003). Moreover, hyperlipidemia is also associated with coronary artery disease, metabolic syndrome, insulin resistance and other associated metabolic diseases (Faraj et al., 2004). Genomic sequencing of the C5L2 coding region in subjects with coronary heart disease identified a naturally occurring C5L2 mutation with a substitution of serine 323 by isoleucine in the C-terminal intracellular region (S323I). This proband was characterized by increased circulating plasma ASP, triglycerides and apolipoprotein B. Increased circulating triglycerides are often associated with delayed plasma clearance due to decreased fatty acid uptake and esterification in storage tissues such as adipose tissue (Cianflone et al., 2003). Although this mutation was not detected in scanning an additional 2176 subjects, 8 out of 17 family members were identified as heterozygote for S323IC5L2. The heterozygote family members had significantly higher plasma ASP, apolipoprotein B and triglyceride levels. Bioactivity and competition binding assays in cells obtained from C5L2 (+/−) subjects showed that ASP stimulation of TG synthesis and glucose transport was reduced by 50%, as was ASP binding, compared to cells obtained from siblings and un-related subjects expressing only wildtype C5L2(+/+), suggesting altered C5L2 function (Marcil et al., 2006). Following ligand activation and internalization, GPCRs are recycled back to cell-surface membrane or targeted for degradation. To study trafficking of internalized receptors, Rab proteins, Ras-like

W. Cui et al. / Molecular Immunology 46 (2009) 3086–3098

small G-proteins, are commonly used. Rab proteins are involved in exocytosis, endocytosis, endosome fusion, and trafficking (Seachrist and Ferguson, 2003). Rab5 is involved in formation and endocytosis of plasma membrane clathrin-coated vesicles and fusion with early endosomes, and subsequent sorting. Rab11 is usually present in perinuclear recycling endosomes and is involved in regulation of slow receptor recycling traffic from early endosomes to recycling endosomes or the trans-golgi network. Rab7 is localized to late endosomes and lysosomes, regulating trafficking of receptors to late endosomes and lysosomes for degradation. In the present study, our aims were: (i) to compare ASP- and C5a-mediated C5L2 internalization using colocalization of receptor with ␤-arrestin-2-GFP, (ii) to study trafficking of internalized C5L2 with Rab5, Rab7, and Rab11 and (iii) to characterize the effects of the previously identified naturally occurring S323I mutation on C5L2 activation and internalization. 2. Materials and methods 2.1. Recombinant ASP and C5a Recombinant ASP (rASP) was purified based on a modification of a previously published method (Murray et al., 1997), using a His-Tag labelled recombinant ASP, with purification using Ni+ affinity chromatography (GE Healthcare, Piscataway, NJ) followed by HPLC (Murray et al., 1997). Fluorescently labelled Fluos-ASP was prepared as described previously (Kalant et al., 2003). Recombinant C5a (rC5a) was purchased from EMD Biosciences (Gibbstown, NJ). Note, as normal human physiological plasma concentrations of ASP range from 25 nM to 100 nM, increasing to 300 nM in obesity (Cianflone et al., 2003), and are therefore 25-fold greater than C5a (1–2 nM)(Huber-Lang et al., 2005), the experimental concentrations used were adjusted accordingly. Further, based on the KD of ASP and C5a (∼50–75 nM and ∼3.5 nM, respectively), in order to obtain rapid maximal effects, concentrations of 10-fold KD which result in >90% receptor occupancy were targeted (Kalant et al., 2003; Johswich et al., 2006; Okinaga et al., 2003). Comparable C5a concentrations were used in other studies (range 10–200 nM) (Kalant et al., 2003; Huber-Lang et al., 2005; Cain and Monk, 2002; Okinaga et al., 2003). 2.2. Cell culture and preparation of stable transfectants HEK 293 cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM)/F-12 medium supplemented with 10% (v/v) fetal bovine serum (FBS) at 37 ◦ C, 5% CO2 . HEK 293 cells were seeded at 1 × 106 cells in 75 cm2 flasks. At 90% confluency, cells were transfected with wildtype hC5L2 containing an N-terminal HA tag (as previously described; Kalant et al., 2003) or mutant hS323I-C5L2 pcDNA3.1(+) plasmid DNA using Lipofectamine 2000 Reagent as described by the manufacturer (Invitrogen, Carlsbad, CA). After 24 h, the transfection solution was removed and replaced by DMEM/F-12 with 10% FBS. One day later, the transfected cells were passaged by 1:4 dilution and the cells were selectively grown with 0.2 ␮g/mL zeocin. Before use, all cell lines were preincubated in serum-free DMEM/F12 media for at least 2 h, to remove the effect of any potential serum stimulation. 2.3. Construction of mutant human C5L2 S323I recombinant pcDNA3.1(+) plasmid Mutant C5L2 receptor cDNAs were constructed by PCR-based site-directed mutagenesis using hC5L2 cDNA pcDNA3.1(+) plasmid as template (previously described in Kalant et al., 2005 (not same plasmid)). Oligonucleotide T7: 5 -TAATACGACTCACTATAGGG3 was used as forward primer. Oligonucleotide 5 -CCC GAA TTC

3089

CTA CAC CTC CAT CTC CGA GAC CAG GTC ATG GCT GGT GGA TTT CTT GAT GTC CAC ACT TTC GTC CTG GCC-3 (Alpha DNA, Montreal, Canada) was used as reverse primer. The mutant nucleotide residue is shown as bold and underlined. PCR reaction was 94 ◦ C (1 min), 55 ◦ C (1 min), 72 ◦ C (1 min) for 30 cycles. After purification (Microcon-PCR centrifugal Filter Devices, Millipore Etobicoke, ON), the PCR product was inserted into pcDNA3.1(+) using HindIII and EcoRI restriction sites. S323I-C5L2 recombinant plasmid was verified by DNA sequencing (Genome Quebec Innovation Centre, McGill University, Montreal, Canada). 2.4. Detection of receptor expression and FACS cell sorting S323I-HEK, C5L2-HEK, and HEK 293 cells were grown to 85% confluency in T-75 flasks. For C5L2-HEK cells, after incubation in serum-free DMEM medium for 2 h, cells were incubated with Fluos-ASP (25 nM) for 30 min at 37 ◦ C. Cells were detached 0.25% trypsin/0.02% EDTA in PBS (Sigma Chemicals, St. Louis, MO) and pelleted. C5L2-HEK cells were further selected by two rounds of Facscan sorting collecting the top 5% fluorescent cells (Kalant et al., 2003). Facscan analysis demonstrated similar fluorescent profiles using either Fluos-ASP or anti-HA antibody. For S323I-HEK, cells were sorted based on receptor expression. After incubation in serum-free DMEM medium for 2 h, cells were detached using a non-enzymatic cell dissociation solution (Sigma Chemicals, St. Louis, MO) and pelleted. The pellet was resuspended in 1 ml PBS with 0.5% BSA containing rabbit anti-HA antibodies (diluted 1:200) (Bethyl, Montgomery, TX), and incubated at 4 ◦ C for 30 min. Cells were washed, incubated with FITC conjugated goat anti-rabbit antibodies (dilution 1:400) (Bethyl, Montgomery, TX) in PBS–BSA (4 ◦ C, 30 min), washed and sorted. For Facscan analysis (without sorting), cells were fixed in 4% paraformaldehyde (PFA) in PBS, which was then diluted to 0.4%. Sorting and FACS were performed at Laboratoire de cytométrie, Université Laval (Quebec, Canada). 2.5. Receptor endocytosis assay C5L2-HEK cells were incubated in serum-free media for 2 h. To block de novo receptor synthesis, cells were preincubated for 30 min in 10 ␮g/ml cycloheximide (which was maintained throughout the experiment) in serum-free media. Cells were stimulated with rASP (final concentration 600 nM) or rC5a (50 nM) for 0.5, 1, 2, 4 h. In some cases, following initial serum-free preincubation, cell media was supplemented with 0.45 M sucrose or 2% hydroxypropyl-␤cyclodextrin for 30 min followed by stimulation with or without rASP or rC5a for 1 h. Cells were collected and prepared for FACS analysis as described above. 2.6. Immunofluorescent confocal microscopy C5L2-HEK and S323I-HEK cells were seeded onto glass coverslips set in 6-well plates and transfected with 0.5 ␮g of ␤arrestin-2-GFP (green fluorescent protein) pEGFP plasmid (Oakley et al., 1999), Rab5-GFP, Rab7-GFP, or Rab11-GFP pEGFP vector (Hunyady et al., 2002) using Lipofectamine 2000 Reagent (as described by the manufacturer, Invitrogen). At 48 h posttransfection, cells were preincubated in serum-free DMEM/F12 media for 2 h, followed by ligand stimulation with 600 nM rASP or 50 nM rC5a for the indicated times. Cells were: (i) fixed in 4% (w/v) PFA–PBS for 8 min, (ii) treated with blocking solution (1% BSA–PBS and 0.2% Triton) for 15 min, followed by rabbit anti-HA antibodies (diluted 1:400) (Bethyl, Montgomery, TX) in blocking solution for 1.5 h at room temperature. After washing, cells were incubated with goat anti-rabbit antibody conjugated with TRITC (dilution 1:1000) (Sigma, Oakville, ON) at room temperature for 1 h. After washing, coverslips were mounted on microscope slides

3090

W. Cui et al. / Molecular Immunology 46 (2009) 3086–3098

using GelTol Mounting Medium (Thermo Electron Corp., Waltham, MA) and dried overnight in the dark. Data were collected using an Olympus FluoView 300 confocal microscope (Center Valley, PA). 2.7. Competition binding assay Competition binding assays were performed using adherent C5L2-HEK, S323I-HEK, and HEK cells at 95% confluence in 24-well cell culture plates. Cells were preincubated for 2 h in serum-free media. The adherent cells were then incubated with 1 nM 125 I-ASP (specific activity = 171 cpm/pmole) in PBS (200 ␮l) with increasing concentrations of unlabeled ASP for 1 h at room temperature. After washing, cells were dissolved in 0.1N NaOH, aliquots counted and cell protein measured by Bradford protein assay (BioRad, Mississauga, ON). Note, in the absence of cells, plates blocked overnight with 10% FBS in DMEM/F12 demonstrated only background binding of 125 I-ASP (data not shown). 2.8. TG synthesis assay and glucose transport assay Triglyceride synthesis and glucose transport were performed as previously described in detail (Kalant et al., 2005). Briefly, cells were prestimulated with ASP (2 h) then triglyceride synthesis was measured as pmol of [3 H] oleate incorporated into triglyceride per mg of soluble cell protein for an additional 2 h. For glucose transport, following prestimulation with ASP for 2 h, glucose transport was measured as pmol uptake of [3 H] 2-deoxyglucose per mg of soluble cell protein over 10 min. 2.9. Receptor phosphorylation assay Analysis of C5L2 receptor phosphorylation was performed as previously described in detail (Kalant et al., 2005). Briefly, following preincubation in serum and phosphate-free medium, cells were incubated with 32 P media for 3 h to prelabel cells, then stimulated with rASP (1 ␮M, 15 min). Cells were lysed, receptor immunoprecipitated, separated by gel electrophoresis and analyzed for total C5L2 (Western blot) and phosphorylated C5L2 (autoradiography). 2.10. Statistical analyses Results were analyzed by ANOVA followed by post hoc tests or by non-linear regression analysis (competition binding) as indicated in the text. Statistical significance was set at p value <0.05, where p ns indicates not significant. 3. Results 3.1. rASP induces ˇ-arrestin-2-mediated C5L2 internalization Previously, we demonstrated that ASP stimulation of C5L2-HEK cells activated ␤-arrestin-2 translocation to the cell membrane followed by formation of endosomes (Kalant et al., 2005), while non-transfected HEK cells were non-responsive. Further, in the absence of ligand, there was no change in ␤-arrestin-2 distribution (Kalant et al., 2005). To demonstrate: (i) that ASP indeed activates C5L2 and (ii) that C5L2 is subsequently internalized we evaluated ␤-arrestin-2-GFP translocation, C5L2 internalization, and joint ␤arrestin-2 and C5L2 colocalization using confocal imagery over an extended time period. As shown in Fig. 1A, in unstimulated cells, C5L2 (TRITC red fluorescence) was localized exclusively to the cellsurface, while ␤-arrestin-2-GFP (green fluorescence) was diffused in the cytosol. Note, before use, all cell lines were preincubated in serum-free DMEM/F12 media for at least 2 h, to remove the effect of any potential serum stimulation. After 5 min stimulation with rASP, ␤-arrestin-2 was activated and translocated to the membrane in a

Fig. 2. rASP and rC5a induce clathrin- and cholesterol-dependent mediated endocytosis of C5L2 with recycling back to the cell-surface membrane. C5L2-HEK cells were preincubated in serum-free medium for 2 h, followed by 0.5 h incubation with10 ␮g/ml cycloheximide in serum-free media with or without 0.45 M sucrose or 2% hydroxypropyl-␤-cyclodextrin. The cells were then stimulated with or without 600 nM rASP (A) or 50 nM C5a (B) for 1 h. For recycling (C), cells were stimulated with or without 600 nM ASP or 50 nM C5a for 0.5, 1, 2, 4 h. For all experiments, cells were detached using a non-enzymatic cell dissociation solution and collected. Cell pellet was resuspended in 1 ml PBS with 0.5% BSA containing rabbit anti-HA antibodies, and incubated at 4 ◦ C for 30 min. Cells were washed, incubated with FITC conjugated goat anti-rabbit antibodies in PBS–BSA at 4 ◦ C for 30 min. After washing, cells were fixed with 4% PFA in PBS for FACS scanning. Values are expressed as percentage of basal C5L2 expression on membrane (where basal = 100%), where **p < 0.01 and ***p < 0.001 by ANOVA analysis.

punctuate pattern, as was the receptor C5L2. At 15 min, both C5L2 and ␤-arrestin-2 were detected intracellularly and at the plasma membrane with colocalization. At 60 min, there was still colocalization, with larger endocytotic vesicles formed. At 90 min, the degree of colocalization was decreasing, with distinct ␤-arrestin-2 free C5L2 present intracellularly. 3.2. rC5a induces a similar profile of ˇ-arrestin-2-mediated C5L2 internalization Previously, we showed that C5a stimulated ␤-arrestin-2 translocation in C5L2 transfected HEK cells in a time frame similar to that of ASP (Kalant et al., 2005), which suggests that C5L2 does

W. Cui et al. / Molecular Immunology 46 (2009) 3086–3098

undergo an internalization process even following interaction with C5a, as well as ASP. As shown in Fig. 1B, at 5 min stimulation with rC5a, ␤-arrestin-2 was still diffused in the cytoplasm, with C5L2 present at the membrane. However, by 15 min, ␤-arrestin-2 was translocated and distributed in a punctuate fashion close to the cell membrane. C5L2 was also localized in a punctuate manner on the cell membrane and colocalized with ␤-arrestin-2, suggesting a potential formation of clathrin-coated pits. By 60 min, most of the fluorescence was localized intracellularly, with little fluorescence remaining at the membrane, the endocytotic colocalization becoming more pronounced over time, with some close to the nucleus. At 90 min, the endosomes further fused to fewer but larger endosomes, some of which were close to the nucleus, others close to the cell membrane.

3.3. Mechanism of ligand-induced C5L2 endocytosis GPCRs utilize several pathways for endocytosis, including clathrin-dependent and caveolae-dependent (cholesterolmediated) pathways and ␤-arrestin-2 may be involved in both. Cells were preincubated with cycloheximide, to block protein synthesis, with/without sucrose or cyclodextrin to block clathrin-dependent or cholesterol-dependent mediated endocytosis, respectively, and

3091

followed by ligand stimulation for 1 h. As shown in Fig. 2A, ASP induced a 36.7% reduction in cell-surface C5L2, while C5a (Fig. 2B) induced a 55.2% reduction. Interestingly, in sucrose and cyclodextrin treated cells, neither ASP nor C5a induced internalization of C5L2. 3.4. Ligand-induced internalization and recycling of C5L2 receptor It has been previously reported that C5L2 did not undergo internalization following C5a stimulation for 10 min, in contrast to rapid C5aR internalization (Cain and Monk, 2002). However Western blot analysis suggests that total C5L2 endogenously expressed in polymorphonuclear neutrophils (PMN) was down-regulated after longer C5a stimulation (Huber-Lang et al., 2005). Our previous study indicated that both ASP and C5a stimulated ␤-arrestin-2 associated endosomes from 15 min on (Kalant et al., 2005). Accordingly the time course of receptor recycling was evaluated. Cells, pretreated with cycloheximide to block de novo protein synthesis, were evaluated for cell-surface C5L2 following stimulation with either rASP or rC5a for up to 4 h. As shown in Fig. 2C, at 0.5 h, membrane surface C5L2 significantly decreased. At 1 h, membrane C5L2 reached its lowest level, followed by a slow return to the cell-surface over the next few hours.

Fig. 3. Rab proteins regulate the trafficking of internalized C5L2. C5L2-HEK cells were seeded onto glass coverslips and transfected with Rab5-GFP (A), Rab7-GFP (B), or Rab11-GFP (C) in pIRES vector. At 48 h post-transfection, cells were preincubated in serum-free media for 2 h, followed by ligand stimulation with 600 nM rASP for 0, 30, 60, and 120 min. After stimulation, cells were: (i) fixed in 4% PFA–PBS for 8 min, (ii) treated with 0.2% Triton in 4% BSA–PBS for 15 min, followed by rabbit anti-HA antibody incubation for 1.5 h at room temperature. After washing, cells were incubated with goat anti-rabbit antibody conjugated with TRITC at room temperature for 1 h. Cells were washed with PBS before the coverslips were mounted on microscope slides. The samples were dried overnight in the dark and visualized using confocal microscopy.

3092

W. Cui et al. / Molecular Immunology 46 (2009) 3086–3098

Fig. 3. (Continued )

3.5. Rab5-mediated trafficking of internalized C5L2 to early endosomes induced by rASP

in perinuclear regions. After 120 min, only a few vesicles containing both Rab7 and C5L2 were still detected.

To evaluate the intracellular pathway of C5L2 after endocytosis, Rab-GFP proteins (Rab5, Rab7, Rab11) were transiently transfected into C5L2-HEK stably transfected cells. The cells were stimulated with rASP at 37 ◦ C for the indicated time. In unstimulated cells, consistent with results shown above, C5L2 was localized primarily to the cell-surface, while Rab5-GFP was diffused in the cytosol and on endosomes (Fig. 3A). At 30 min, partial colocalization of C5L2 was detected with Rab5. By 60 min, colocalization fluorescence was centred in fewer larger endosomes in the perinuclear region. By 120 min, little C5L2 colocalized with Rab5/sorting endosomes. Distinct cell-surface C5L2 could again be identified.

3.7. Rab11-regulated recycling of internalized C5L2

3.6. Rab7-mediated degradation of internalized C5L2 induced by rASP Rab7, involved in late endosomes and lysosomal trafficking, is shown in Fig. 3B. Rab7 was distributed in cytosol and endosomes, while C5L2 was found at the cell-surface in unstimulated cells. At 30 min stimulation with rASP, there was prominent redistribution of C5L2. Although Rab7 was detected in scattered cytosolic vesicles, C5L2 was now found in a punctate pattern, with numerous colocalization areas, suggesting that Rab7 is involved in the intracellular process of trafficking C5L2 to late endosomes and/or lysosomes. At 60 min, the scattered small vesicles had fused into large endosomes

As demonstrated above (Fig. 2C), internalized C5L2 recycled back to the membrane, suggesting involvement of recycling Rab protein, such as Rab4 or Rab11. Accordingly, Rab4 or Rab11 was transiently transfected into stably expressing C5L2-HEK cells, followed by treatment with rASP for the indicated times. Confocal microscopy analysis showed that, in unstimulated cells, Rab11 appeared in dense perinuclear patches. At 30 min stimulation, no change was found in the cell distribution of Rab11 proteins although now C5L2 strongly colocalized with Rab11 (as indicated by yellow in merge). By 60 min, the proteins were found in small cytoplasmic vesicles. By 120 min there was no colocalization, and the pattern had returned to the original state of unstimulated cells (Rab11 perinuclear, C5L2 plasma membrane). For Rab4, throughout the time period of stimulation, there was no colocalization with C5L2 suggesting that this protein may not be involved in C5L2 trafficking (data not shown). 3.8. Preparation of C5L2 mutant expressing cell line Previously, we identified a naturally occurring C5L2 mutation (S323I), where cells heterozygote for the mutation had reduced response to ASP stimulation of TG synthesis and glucose transport

W. Cui et al. / Molecular Immunology 46 (2009) 3086–3098

3093

Fig. 3. (Continued ).

(Marcil et al., 2006). To investigate the activation and response of this receptor, in comparison to native C5L2, a S323I-C5L2 containing an N-terminal HA tag was constructed by site-directed mutagenesis, into a plasmid, and stably transfected into HEK cells. The cells were sorted using anti-HA antibody recognition and the top 5% of S323I expressing cells were collected. FACS analysis showed that cell-surface S323I was expressed in the transfected cell line, at a lower level compared to the wildtype C5L2 (Fig. 4A), in spite of multiple sorting. To evaluate binding and uptake of ASP in S323IHEK, Fluos-ASP was used. FACS results showed that wildtype C5L2 expressing cells bound and took up much more than S323I expressing cells (Fig. 4B). Further study on competition binding showed

that ASP bound to S323I with similar affinity as wildtype C5L2 (EC50 : C5L2 3.1 × 10−7 M and S323I 6.3 × 10−7 M), but with fewer cell-surface binding sites (Fig. 4C). 3.9. S323I-C5L2 loses ASP-induced C5L2 phosphorylation ASP (Kalant et al., 2005) and C5a (Okinaga et al., 2003) treatment of C5L2 transfected cells results in time-dependent increases in phosphorylation of C5L2. The S323I mutation, which is localized to the C-terminal intracellular region of C5L2 in a Ser–Thr rich region (Marcil et al., 2006) is a prime phosphorylation site. The effects of serine to isoleucine mutation on C5L2 phosphorylation

Table 1 Effects of S323I-C5L2 mutant on triglyceride synthesis and glucose transport. [ASP] (␮M)

HEK S323I-HEK C5L2-HEK

Triglyceride synthesis

Glucose transport

0

2.5

5

10

0

100.0 ± 2.2 100.0 ± 17.4 100.03 ± 1.0

129.6 ± 53.1 74.1 ± 7.2 338.5 ± 10.0*

118.2 ± 16.3 124.5 ± 30.0 289.6 ± 22.0*

111.8 ± 29.3 102.9 ± 27.0 307.9 ± 33.0*

95.6 ± 2.2 97.0 ± 7.4 96.3 ± 2.9

2.5 76.9 ± 5.5 90.2 ± 5.6 155.1 ± 21.0*

5

10

84.4 ± 14.0 109.0 ± 2.7 180.5 ± 7.4*

95.1 ± 9.9 115.7 ± 10.9 179.0 ± 20.0*

After preincubation in SF media for 2 h, HEK, C5L2-HEK, and S323I-HEK were incubated with ASP at the indicated concentrations for 2 h, and then 100 ␮M [3 H] oleate was added and incubated for a further 2 h at 37 ◦ C. After washing, lipids were extracted, separated by thin layer chromatography and counted. [3 H] oleate incorporation into triglyceride is measured as pmol of [3 H] TG/mg of cell protein. For glucose transport, following incubation with ASP at the indicated concentrations, cells were rinsed with warm serum-free, glucose-free media and then incubated for 5 min with [3 H] 2-deoxyglucose in serum-free glucose-free media at 37 ◦ C. Cells were dissolved, aliquots counted and protein measured. Glucose transport was measured as pmol of [3 H] 2-deoxyglucose uptake per mg of soluble cell protein. Values are expressed as percentage of basal TG synthesis or glucose transport. Results are presented as mean ± SEM for n = 6 for each value. * p < 0.05 by ANOVA analysis.

3094

W. Cui et al. / Molecular Immunology 46 (2009) 3086–3098

3.10. S323I-C5L2 is unresponsive to ASP stimulation As shown in Table 1, ASP stimulates triglyceride synthesis and glucose transport in a concentration dependent manner in stable transfected C5L2-HEK cells, while non-transfected HEK cells are non-responsive. By contrast, S323I-C5L2 transfected HEK cells showed that neither triglyceride synthesis nor glucose transport was stimulated at any concentration of ASP. 3.11. S323I-C5L2 mutant loses ligand-induced internalization While ASP appears to bind to S323I-C5L2, both phosphorylation and functional activity (triglyceride synthesis and glucose transport stimulation) are absent, suggesting a potential defect in signalling and trafficking. This suggests that C5L2 may not undergo ␤-arrestin2-mediated endocytosis. To test the hypothesis, S323I-C5L2 stably transfected HEK 293 cells were transiently transfected with ␤arrestin-2-GFP. The cells were incubated with rASP or rC5a for the indicated times, respectively. As shown in Fig. 5A, ligand stimulation of S323I-C5L2 did not induce ␤-arrestin-2 translocation to the cell membrane, even at activation times with rASP up to 60 min. Similarly, ligand activation with rC5a did not stimulate ␤-arrestin2 translocation at any time point tested from 15 to 120 min (data not shown). In addition, S323I-C5L2 mutant was detected both on the plasma membrane and intracellularly even in unstimulated cells. 3.12. S323I-C5L2 mutant has altered receptor distribution on membrane

Fig. 4. Expression, binding and phosphorylation assay of S323I-C5L2 stably transfected mutant in HEK 293 cells. HEK, S323I-HEK, and C5L2-HEK cells were preincubated in serum-free medium for 2 h. (A) Cells were detached using a nonenzymatic cell dissociation solution and collected. Cells were resuspended in 1 ml PBS with 0.5% BSA containing rabbit anti-HA antibodies, and incubated at 4 ◦ C for 30 min. Cells were washed, incubated with FITC conjugated goat anti-rabbit antibodies in PBS–BSA (4 ◦ C, 30 min). After washing, cells were fixed with 4% PFA in PBS for Facscan. (B) After preincubation, cells were incubated with Fluos-labeled ASP for 30 min at 37 ◦ C and washed three times. Cells were detached with 0.25% trypsin containing 0.02% EDTA in PBS, fixed with 1% PFA, washed with PBS for Facscan. (C) HEK-C5L2, S323I-HEK, HEK cells were grown in 24-well cell culture plates. Cells were preincubated for 2 h in serum-free media. The adherent cells were then incubated with 1 nM 125 I-ASP in PBS with increasing concentrations of unlabeled ASP for 1 h at room temperature. After washing, cells were dissolved in 0.1N NaOH, aliquots counted and cell protein measured by Bradford protein assay. In the absence of cells (plates preincubated overnight in media with FBS), ASP binding was at background levels only. Values are presented as mean ± SEM. (D) C5L2-HEK and S323I-HEK cells were preincubated in serum-free phosphate-free medium for 4 h. After washing, cells were incubated for 2 h with 32 P followed by stimulating with or without ASP for 15 min. After lysing, cell debris was pelleted. C5L2 and C5L2 mutant were immunoprecipitated with anti-HA-agarose beads. After three washes, samples were incubated for 30 min at 65 ◦ C in sample buffer (containing SDS and ␤-mercaptoethanol), electrophoresed on SDS-PAGE, and autoradiographed.

were tested by immunoprecipitation of C5L2 from wildtype C5L2 and S323I stably transfected HEK 293 cells that had been metabolically labelled with 32 P. As shown in Fig. 4D, following stimulation with rASP, there is a strong induction of C5L2 phosphorylation in wildtype C5L2 cells. By contrast, ASP binding to S323I did not stimulate S323I-C5L2 phosphorylation. Based on Western blot analysis, total cellular C5L2 was comparable in C5L2 vs. S323I mutated C5L2 cells, although this does not differentiate between cell-surface vs. intracellular protein.

In the current study, we found an interesting phenomenon, that S323I-C5L2 cell-surface expression in transfected cells could not be maintained at a high level, in spite of FACS selection for highly expressing cells. Following multiple rounds of growth, the cells reverted to lower cell-surface expression, in spite of the presence of maintenance antibiotic. In addition, competition binding studies demonstrated a lower total binding capacity than C5L2-HEK cells. This suggested a reduced capacity for maintenance of the receptor at the cell-surface, in spite of lack of ligand-mediated endocytosis. Accordingly we examined the basal expression level of S323I-C5L2 and wildtype C5L2 in the presence and absence of cell permeabilization, followed by anti-HA visualization. Confocal microscopy analysis showed that the level of S323I-C5L2 mutant on the cellsurface membrane is significantly lower than the wildtype C5L2 (Fig. 5B), relative to the amount present intracellularly. 4. Discussion In the present study we address directly two controversial issues: (i) is ASP/C3adesArg a ligand for C5L2? and (ii) is C5L2 a receptor that is activated, internalized and recycled following ligand stimulation? Our data suggest that both ASP and C5a bind and activate C5L2, and that ASP stimulates a functional response. Further support for this is demonstrated by the lack of activation by C5a and ASP of the C5L2 S323I mutation. Specifically, we clearly demonstrate that with both C5a and ASP/C3adesArg treatment, C5L2 is phosphorylated, ␤-arrestin2 is translocated to the plasma membrane, and that receptors are internalized and colocalized with ␤-arrestin-2. This internalization is potentially clathrin- and cholesterol-dependent, and leads to transient colocalization in endocytotic vesicles with Rab5, Rab7 and Rab11, followed by partial recycling back to the plasma membrane. By contrast, a S323I-C5L2 mutation in the Ser–Thr rich C-terminal region is defective in ligand-dependent phosphorylation, ␤-arrestin-2 activation and colocalization, and cell-surface presentation is altered. This mutation leads to an

W. Cui et al. / Molecular Immunology 46 (2009) 3086–3098

absence of ASP stimulation of triglyceride synthesis and glucose transport, in contrast to wildtype C5L2 which is functionally active. With respect to ligand binding, while we and Monk in collaborative studies have independently demonstrated ASP/C3adesArg ligand activity (Kalant et al., 2003, 2005), Gerard and colleagues (Honczarenko et al., 2005) and Klos and colleagues (Johswich et al., 2006) have not. Indeed, Klos has suggested that their apparent binding was primarily related to non-specific C3a binding to BSA, however, in our present competition binding we do not use BSA for blocking, and therefore did not encounter this problem. There are several potential possibilities to explain the differences. Commercially available C3a/C3adesArg is usually prepared using denaturants. This has been shown to inactivate binding and activity of ASP/C3adesArg (Murray et al., 1999), although denaturation, which involves unfolding of 3 cysteine bridges, has no effect on C3a activity or interaction with the C3a receptor (Cain and Monk, 2002). Accordingly, both our plasma ASP and recombinant ASP preparations are entirely without the use of denaturants (Murray et al., 1997). Further, recombinant ASP has been used in the present studies to prevent any effect that could be ascribed to any minor

3095

contaminant derived from plasma. Future studies using this available recombinant ASP will be able to resolve these issues. With respect to C5L2 activation, in the current study, we clearly demonstrated that not only ASP but also C5a stimulated time-dependent ␤-arrestin-2 translocation followed by C5L2 internalization and colocalization with ␤-arrestin-2 and Rab proteins. Extended analysis demonstrated C5L2 recycling and return to the plasma membrane. Furthermore, it was demonstrated that treatment with ASP (in this study and previously in Kalant et al., 2005) resulted in time-dependent phosphorylation of C5L2. Okinaga demonstrated C5a-mediated C5L2 phosphorylation beginning at 5 min, reaching a maximum at 10 min, and decreasing thereafter although there was still some phosphorylation remaining at 15 min (Okinaga et al., 2003). Okinaga et al. (2003) and Cain and Monk (2002) showed that C5a did not induce C5L2 internalization following 5 to 10 min stimulation. However, Huber-Lang et al. (2005) demonstrated effective decrease in C5L2 protein after 3 h of C5a treatment of PMN cells. This is consistent with our present results that demonstrate internalization of C5L2 is only beginning to be detectable after 15 min, reaching a maximum around 1 h. Clearly the C5L2 cell lifecycle is substantially slower than that of the C5aR

Fig. 5. ASP binding to S323I-C5L2 mutation did not activate ␤-arrestin-2 translocation, and S323I-C5L2 receptors are present intracellularly. (A) S323I-HEK cells were seeded onto glass coverslips and transfected with ␤-arrestin-2-GFP pEGFP plasmid. At 48 h post-transfection, cells were preincubated in serum-free media for 2 h, followed by ligand stimulation with 600 nM rASP. After stimulation, cells were: (i) fixed in 4% PFA–PBS for 8 min, (ii) permealized with 0.2% Triton 1% in BSA–PBS for 15 min, followed by rabbit anti-HA antibody incubation for 1.5 h at room temperature. After washing, cells were incubated with goat anti-rabbit antibody conjugated with TRITC at room temperature for 1 h. Cells were washed with PBS before the coverslips were mounted on microscope slides. The samples were dried overnight in the dark, and visualized using confocal microscopy. (B) S323I-HEK and C5L2-HEK cells were seeded onto glass coverslips and preincubated in serum-free media for 2 h. Cells were: (i) fixed in 4% PFA–PBS for 8 min, (ii) treated with 1% BSA–PBS with/without 0.2% Triton for 15 min, followed by rabbit anti-HA antibody incubation for 1.5 h at room temperature. After washing, cells were incubated with goat anti-rabbit antibody conjugated with TRITC at room temperature for 1 h. Cells were washed with PBS before the coverslips were mounted on microscope slides. The samples were dried overnight in the dark and visualized using confocal microscopy.

3096

W. Cui et al. / Molecular Immunology 46 (2009) 3086–3098

Fig. 5. (Continued ).

which internalizes and recycles much more quickly (Van Epps et al., 1990). During preparation of this manuscript, a recent detailed study by Scola et al. (2008) demonstrated clathrin-mediated C5L2 internalization and recycling back to the plasma membrane, with associated ligand (C5a) degradation. This was demonstrated in both transfected RBL cells, as well as in cells endogenously expressing C5L2 at comparable, higher or lower levels than C5a receptor. Overall, using different methodology and cells, their study and the present study are reasonably consistent in results, although not in final interpretation. Specifically, we demonstrate a primarily cellsurface localization of C5L2 in unstimulated conditions. This may be a consequence of the consistent minimum 2-h preincubation period in serum-free media to dispense of any residual activation caused by hormones in serum or autocrine ligand. This methodology is typically used in metabolic studies, such as with insulin (Liao et al., 2007). This distinct cell-surface localization was not present in the S323I-C5L2 mutation. Further, while we demonstrated ␤-arrestin-2-GFP translocation, Monk and colleagues (Scola et al., 2008) did not identify any translocation of endogenous ␤-arrestin-1. However, it should be noted that ␤-arrestin-1 and ␤-arrestin-2 do not play redundant roles (DeWire et al., 2007), receptors vary in their affinities for ␤-arrestin-1 and ␤-arrestin2 (Oakley et al., 2000), this can be further influenced by the cell types used (Neel et al., 2005), and finally, even C5aR has been suggested to have a higher affinity for ␤-arrestin-2 (Braun et al., 2003). On the other hand, our results dovetail with theirs demonstrating clathrin-mediated C5L2 internalization and recycling back to the plasma membrane. Further, efficient C5a degradation is associated with C5L2 internalization. While it is suggested that this may be evidence of a decoy receptor (Weber et al., 2004), this is typical of hormone receptors, as demonstrated in studies

with the classical hormone insulin (Marshall, 1985; Olefsky et al., 1982). Rab5 protein is involved in regulating the internalized C5L2 trafficking induced by ASP. This provides additional evidence that C5L2 is undergoing clathrin-mediated endocytosis, as Rab5 is involved in formation of clathrin-coated vesicles, endocytosis of the vesicles, and fusion of the vesicles with early endosomes (Markgraf et al., 2007). While Rab4 was not involved in C5L2 trafficking, Rab11 clearly plays a role. This is consistent with the data showing that internalized C5L2 needs at least 1 h to begin recycling back to the membrane, as Rab4 usually controls rapid recycling while Rab11 often regulates slower receptor recycling (Seachrist and Ferguson, 2003; Sheff et al., 1999). Further, Rab7 colocalization with internalized C5L2, suggests that in addition to recycling, C5L2 may also undergo degradation. This raises the question of C5L2 function. Based on the present study, evaluation of C5L2 function may need to be examined on a different time scale or for different functions than those that are typically used in C5aR evaluation. Previous studies have evaluated C5L2 transfected cells or cells endogenously expressing C5L2 for changes in intracellular Ca2+ over short time periods (1–4 min) (Cain and Monk, 2002; Johswich et al., 2006; Okinaga et al., 2003; Scola et al., 2008), at a time when C5L2 phosphorylation, an early step in activation, is not yet even maximal. ASP stimulation of triglyceride synthesis and glucose transport is typically done over longer time periods. Initial studies on ASP activity demonstrated that incubation for at least 30 min is required before an increase in glucose uptake is noted (Germinario et al., 1993), and a similar time frame is required for triglyceride synthesis (Germinario et al., 1993). Studies on S323I-C5L2 mutation demonstrate the consequence of a lack of phosphorylation capacity (in spite of apparent ligand binding), which abrogates ␤-arrestin-2 translocation,

W. Cui et al. / Molecular Immunology 46 (2009) 3086–3098

endocytic colocalization and downstream functional activity for triglyceride synthesis and glucose transport. The functional consequences in transfected HEK cells are reflected by similar decreases in cells derived from patients endogenously expressing the mutation (Marcil et al., 2006). Alternately, changes in Ca2+ may not be involved in the C5L2 pathway; there are many downstream effectors of G-protein activation, many of which can be Ca2+ independent, including those that have been shown to be involved in ASP/C5L2 activation (ERK1/2, Akt, phospholipase A2, protein kinase C) (Brambilla et al., 2002; Chen et al., 2007; Kehrl, 1998; Maslowska et al., 2006). When incubated with C5a alone, there was no stimulation of ␤-hexosaminidase release in C5L2-HEK cells, while C5aR-HEK provided a robust response (Cain and Monk, 2002). However, while C5a alone did not induce an immune response, C5a did potentiate ␤-hexosaminidase release following IgE treatment (Cain and Monk, 2002). The result suggests that C5a-associated C5L2 function may need to be initialized in order to obtain a response, or that the function may be different from previously identified C5aR functions. The absence of Ca2+ changes or hexosaminidase activity is not sufficient to confirm that C5L2 is a decoy receptor. In vivo studies in C5L2 KO mice support a physiological role for C5L2, although again, not without some controversy in results. We have provided evidence of a role for C5L2 in lipid metabolism in mice that are unchallenged or unstressed physiologically. Absence of C5L2 results in delayed fat clearance following a meal, and a slight hyperphagia more than offset by increased fat oxidation, providing an obesity protection (Paglialunga et al., 2008). This is a phenotype comparable to that seen in mice that are deficient in ASP (through lack of the precursors C3 or factor B, required for ASP production) (Cianflone et al., 2003; Paglialunga et al., 2008; Roy et al., 2008). In addition to knockout mice, targeted blocking of C5L2 with antibodies also supports a function role for C5L2. A 10-day injection of an anti-C5L2 neutralizing antibody showed a delayed TG and glucose clearance (Cui et al., 2007). In an immune-challenged situation, Gerard and colleagues (Gerard et al., 2005) demonstrated that targeted deletion of C5L2 resulted in increased neutrophils, IL-6 and TNF␣. In addition, cells from C5L2 knockout (KO) mouse demonstrated increased chemotactic activity compared with wildtype cells after C5a stimulation. Histologic analysis showed enhanced inflammation in C5L2 KO mice. These data suggest that loss of C5L2 increased immune response. On the other hand, Chen et al. (2007) showed that C5L2 KO had reduced inflammatory response. C5L2(−/−) neutrophils produced less Mac-1 in response to C5a, and were no longer responsive to C5a inhibition of TNF-␣ and stimulation of IL-6 production. In addition, histological analyses of OVA-induced lung tissue demonstrated that C5L2 KO mice had significantly reduced allergen-induced inflammatory cell infiltration. Contrary to Chen’s study, Gerard and his colleagues showed that gene deletion of C5L2 resulted in increased mice mortality after an LPS challenge (Gerard et al., 2005). By contrast, Rittirsch et al. (2008) published opposing data demonstrating that C5L2 KO mice had an increased survival ratio compared to wildtype mice following cecal-ligation and puncture treatment. Although these data showed controversial results and need further investigation, nonetheless, they provide evidence of C5L2 functionality. 5. Conclusion In conclusion, the present paper demonstrated that both ASP and C5a are cognate ligands of C5L2. ASP as well as C5a induces ␤-arrestin-2 recruitment, C5L2 endocytosis, and recycling. As ASP is an anabolic hormone and C5a is an inflammatory factor, C5L2 may be one more bridge between the immune and the adipose systems (MacLaren et al., 2008), integrating multiple physiological systems.

3097

Acknowledgements This study was supported by a grant from CIHR (#64446 to KC). K. Cianflone holds a Canada Research Chair in Adipose Tissue. References Baldo, A., Sniderman, A.D., St Luce, S., Zhang, X.J., Cianflone, K., 1995. Signal transduction pathway of acylation stimulating protein: involvement of protein kinase C. J. Lipid Res. 36, 1415–1426. Borroni, E.M., Bonecchi, R., Buracchi, C., Savino, B., Mantovani, A., Locati, M., 2008. Chemokine decoy receptors: new players in reproductive immunology. Immunol. Invest. 37, 483–497. Brambilla, R., Neary, J.T., Cattabeni, F., Cottini, L., D’Ippolito, G., Schiller, P.C., Abbracchio, M.P., 2002. Induction of COX-2 and reactive gliosis by P2Y receptors in rat cortical astrocytes is dependent on ERK1/2 but independent of calcium signalling. J. Neurochem. 83, 1285–1296. Braun, L., Christophe, T., Boulay, F., 2003. Phosphorylation of key serine residues is required for internalization of the complement 5a (C5a) anaphylatoxin receptor via a beta-arrestin, dynamin, and clathrin-dependent pathway. J. Biol. Chem. 278, 4277–4285. Burg, M., Martin, U., Bock, D., Rheinheimer, C., Kohl, J., Bautsch, W., Klos, A., 1996. Differential regulation of the C3a and C5a receptors (CD88) by IFN-gamma and PMA in U937 cells and related myeloblastic cell lines. J. Immunol. 157, 5574– 5581. Cain, S.A., Monk, P.N., 2002. The orphan receptor C5L2 has high affinity binding sites for complement fragments C5a and C5a des-Arg(74). J. Biol. Chem. 277, 7165–7169. Chen, N.J., Mirtsos, C., Suh, D., Lu, Y.C., Lin, W.J., McKerlie, C., Lee, T., Baribault, H., Tian, H., Yeh, W.C., 2007. C5L2 is critical for the biological activities of the anaphylatoxins C5a and C3a. Nature 446, 203–207. Chenoweth, D.E., Goodman, M.G., 1983. The C5a receptor of neutrophils and macrophages. Agents Actions Suppl. 12, 252–273. Cianflone, K., Xia, Z., Chen, L.Y., 2003. Critical review of acylation-stimulating protein physiology in humans and rodents. Biochim. Biophys. Acta 1609, 127–143. Cui, W., Paglialunga, S., Kalant, D., Lu, H., Roy, C., Laplante, M., Deshaies, Y., Cianflone, K., 2007. Acylation-stimulating protein/C5L2-neutralizing antibodies alter triglyceride metabolism in vitro and in vivo. Am. J. Physiol. Endocrinol. Metab. 293, E1482–E1491. DeWire, S.M., Ahn, S., Lefkowitz, R.J., Shenoy, S.K., 2007. Beta-arrestins and cell signaling. Annu. Rev. Physiol. 69, 483–510. Faraj, M., Lu, H.L., Cianflone, K., 2004. Diabetes, lipids, and adipocyte secretagogues. Biochem. Cell Biol. 82, 170–190. Fayyazi, A., Scheel, O., Werfel, T., Schweyer, S., Oppermann, M., Gotze, O., Radzun, H.J., Zwirner, J., 2000. The C5a receptor is expressed in normal renal proximal tubular but not in normal pulmonary or hepatic epithelial cells. Immunology 99, 38–45. Gerard, N.P., Gerard, C., 1991. The chemotactic receptor for human C5a anaphylatoxin. Nature 349, 614–617. Gerard, N.P., Lu, B., Liu, P., Craig, S., Fujiwara, Y., Okinaga, S., Gerard, C., 2005. An antiinflammatory function for the complement anaphylatoxin C5a-binding protein, C5L2. J. Biol. Chem. 280, 39677–39680. Germinario, R., Sniderman, A.D., Manuel, S., Lefebvre, S.P., Baldo, A., Cianflone, K., 1993. Coordinate regulation of triacylglycerol synthesis and glucose transport by acylation-stimulating protein. Metabolism 42, 574–580. Giannini, E., Boulay, F., 1995. Phosphorylation, dephosphorylation, and recycling of the C5a receptor in differentiated HL60 cells. J. Immunol. 154, 4055– 4064. Honczarenko, M., Lu, B., Nicholson-Weller, A., Gerard, N.P., Silberstein, L.E., Gerard, C., 2005. C5L2 receptor is not involved in C3a/C3a-desArg-mediated enhancement of bone marrow hematopoietic cell migration to CXCL12. Leukemia 19, 1682–1683. Huber-Lang, M., Sarma, J.V., Rittirsch, D., Schreiber, H., Weiss, M., Flierl, M., Younkin, E., Schneider, M., Suger-Wiedeck, H., Gebhard, F., McClintock, S.D., Neff, T., Zetoune, F., Bruckner, U., Guo, R.F., Monk, P.N., Ward, P.A., 2005. Changes in the novel orphan, C5a receptor (C5L2), during experimental sepsis and sepsis in humans. J. Immunol. 174, 1104–1110. Hunyady, L., Baukal, A.J., Gaborik, Z., Olivares-Reyes, J.A., Bor, M., Szaszak, M., Lodge, R., Catt, K.J., Balla, T., 2002. Differential PI 3-kinase dependence of early and late phases of recycling of the internalized AT1 angiotensin receptor. J. Cell Biol. 157, 1211–1222. Johswich, K., Martin, M., Thalmann, J., Rheinheimer, C., Monk, P.N., Klos, A., 2006. Ligand specificity of the anaphylatoxin C5L2 receptor and its regulation on myeloid and epithelial cell lines. J. Biol. Chem. 281, 39088–39095. Kalant, D., Cain, S.A., Maslowska, M., Sniderman, A.D., Cianflone, K., Monk, P.N., 2003. The chemoattractant receptor-like protein C5L2 binds the C3a desArg77/acylation-stimulating protein. J. Biol. Chem. 278, 11123–11129. Kalant, D., MacLaren, R., Cui, W., Samanta, R., Monk, P.N., Laporte, S.A., Cianflone, K., 2005. C5L2 is a functional receptor for acylation-stimulating protein. J. Biol. Chem. 280, 23936–23944. Kehrl, J.H., 1998. Heterotrimeric G protein signaling: roles in immune function and fine-tuning by RGS proteins. Immunity 8, 1–10. Kohl, J., 2001. Anaphylatoxins and infectious and non-infectious inflammatory diseases. Mol. Immunol. 38, 175–187.

3098

W. Cui et al. / Molecular Immunology 46 (2009) 3086–3098

Lee, D.K., George, S.R., Cheng, R., Nguyen, T., Liu, Y., Brown, M., Lynch, K.R., O’Dowd, B.F., 2001. Identification of four novel human G protein-coupled receptors expressed in the brain. Brain Res. Mol. Brain Res. 86, 13–22. Liao, W., Nguyen, M.T., Yoshizaki, T., Favelyukis, S., Patsouris, D., Imamura, T., Verma, I.M., Olefsky, J.M., 2007. Suppression of PPAR-gamma attenuates insulinstimulated glucose uptake by affecting both GLUT1 and GLUT4 in 3T3-L1 adipocytes. Am. J. Physiol. Endocrinol. Metab. 293, E219–E227. MacLaren, R., Cui, W., Cianflone, K., 2008. Adipokines and the immune system: an adipocentric view. Adv. Exp. Med. Biol. 632, 1–21. Marcil, M., Vu, H., Cui, W., Dastani, Z., Engert, J.C., Gaudet, D., Castro-Cabezas, M., Sniderman, A.D., Genest Jr., J., Cianflone, K., 2006. Identification of a novel C5L2 variant (S323I) in a French Canadian family with familial combined hyperlipemia. Arterioscler. Thromb. Vasc. Biol. 26, 1619–1625. Markgraf, D.F., Peplowska, K., Ungermann, C., 2007. Rab cascades and tethering factors in the endomembrane system. FEBS Lett. 581, 2125–2130. Marshall, S., 1985. Degradative processing of internalized insulin in isolated adipocytes. J. Biol. Chem. 260, 13517–13523. Maslowska, M., Legakis, H., Assadi, F., Cianflone, K., 2006. Targeting the signaling pathway of acylation stimulating protein. J. Lipid Res. 47, 643–652. Murray, I., Kohl, J., Cianflone, K., 1999. Acylation-stimulating protein (ASP): structurefunction determinants of cell surface binding and triacylglycerol synthetic activity. Biochem. J. 342 (Pt 1), 41–48. Murray, I., Parker, R.A., Kirchgessner, T.G., Tran, J., Zhang, Z.J., Westerlund, J., Cianflone, K., 1997. Functional bioactive recombinant acylation stimulating protein is distinct from C3a anaphylatoxin. J. Lipid Res. 38, 2492–2501. Naik, N., Giannini, E., Brouchon, L., Boulay, F., 1997. Internalization and recycling of the C5a anaphylatoxin receptor: evidence that the agonist-mediated internalization is modulated by phosphorylation of the C-terminal domain. J. Cell Sci. 110 (Pt 19), 2381–2390. Neel, N.F., Schutyser, E., Sai, J., Fan, G.H., Richmond, A., 2005. Chemokine receptor internalization and intracellular trafficking. Cytokine Growth Factor Rev. 16, 637–658. Oakley, R.H., Laporte, S.A., Holt, J.A., Barak, L.S., Caron, M.G., 1999. Association of beta-arrestin with G protein-coupled receptors during clathrin-mediated endocytosis dictates the profile of receptor resensitization. J. Biol. Chem. 274, 32248– 32257. Oakley, R.H., Laporte, S.A., Holt, J.A., Caron, M.G., Barak, L.S., 2000. Differential affinities of visual arrestin, beta arrestin1, and beta arrestin2 for G proteincoupled receptors delineate two major classes of receptors. J. Biol. Chem. 275, 17201–17210.

Ohno, M., Hirata, T., Enomoto, M., Araki, T., Ishimaru, H., Takahashi, T.A., 2000. A putative chemoattractant receptor, C5L2, is expressed in granulocyte and immature dendritic cells, but not in mature dendritic cells. Mol. Immunol. 37, 407–412. Okinaga, S., Slattery, D., Humbles, A., Zsengeller, Z., Morteau, O., Kinrade, M.B., Brodbeck, R.M., Krause, J.E., Choe, H.R., Gerard, N.P., Gerard, C., 2003. C5L2, a nonsignaling C5A binding protein. Biochemistry 42, 9406–9415. Olefsky, J.M., Marshall, S., Berhanu, P., Saekow, M., Heidenreich, K., Green, A., 1982. Internalization and intracellular processing of insulin and insulin receptors in adipocytes. Metabolism 31, 670–690. Paglialunga, S., Fisette, A., Yan, Y., Deshaies, Y., Brouillette, J.F., Pekna, M., Cianflone, K., 2008. Acylation-stimulating protein deficiency and altered adipose tissue in alternative complement pathway knockout mice. Am. J. Physiol. Endocrinol. Metab. 294, E521–E529. Paglialunga, S., Schrauwen, P., Roy, C., Moonen-Kornips, E., Lu, H., Hesselink, M.K., Deshaies, Y., Richard, D., Cianflone, K., 2007. Reduced adipose tissue triglyceride synthesis and increased muscle fatty acid oxidation in C5L2 knockout mice. J. Endocrinol. 194, 293–304. Perez, H.D., 1984. Biologically active complement (C5)-derived peptides and their relevance to disease. Crit. Rev. Oncol. Hematol. 1, 199–225. Rittirsch, D., Flierl, M.A., Nadeau, B.A., Day, D.E., Huber-Lang, M., Mackay, C.R., Zetoune, F.S., Gerard, N.P., Cianflone, K., Kohl, J., Gerard, C., Sarma, J.V., Ward, P.A., 2008. Functional roles for C5a receptors in sepsis. Nat. Med. 14, 551–557. Roy, C., Paglialunga, S., Fisette, A., Schrauwen, P., Moonen-Kornips, E., St Onge, J., Hesselink, M.K., Richard, D., Joanisse, D.R., Cianflone, K., 2008. Shift in metabolic fuel in acylation-stimulating protein-deficient mice following a high-fat diet. Am. J. Physiol. Endocrinol. Metab. 294, E1051–E1059. Scola, A.M., Johswich, K.O., Morgan, B.P., Klos, A., Monk, P.N., 2008. The human complement fragment receptor, C5L2, is a recycling decoy receptor. Mol. Immunol.. Seachrist, J.L., Ferguson, S.S., 2003. Regulation of G protein-coupled receptor endocytosis and trafficking by Rab GTPases. Life Sci. 74, 225–235. Sheff, D.R., Daro, E.A., Hull, M., Mellman, I., 1999. The receptor recycling pathway contains two distinct populations of early endosomes with different sorting functions. J. Cell Biol. 145, 123–139. Van Epps, D.E., Simpson, S., Bender, J.G., Chenoweth, D.E., 1990. Regulation of C5a and formyl peptide receptor expression on human polymorphonuclear leukocytes. J. Immunol. 144, 1062–1068. Weber, M., Blair, E., Simpson, C.V., O’Hara, M., Blackburn, P.E., Rot, A., Graham, G.J., Nibbs, R.J., 2004. The chemokine receptor D6 constitutively traffics to and from the cell surface to internalize and degrade chemokines. Mol. Biol. Cell 15, 2492–2508.