The pro-inflammatory chemokine CCL2 [chemokine (Cys-Cys motif) ligand 2; also known as MCP-1 (monocyte chemotactic protein-1)] is up-regulated in the glomerular compartment during the early phase of LPS (lipopolysaccharide)-induced nephritis. This up-regulation also occurs in cultured MCs (mesangial cells) and is more pronounced in MCs lacking the PGE2 (prostaglandin E2) receptor EP2 or in MCs treated with a prostaglandin EP4 receptor antagonist. To examine a possible feedback mechanism of EP receptor stimulation on CCL2 expression, we used an in vitro model of MCs with down-regulated EP receptor expression. Selectively overexpressing the various EP receptors in these cells then allows the effects on the LPS-induced CCL2 expression to be examined. Cells were stimulated with LPS and CCL2 gene expression was examined and compared with LPS-stimulated, mock-transfected PTGS2 [prostaglandin-endoperoxide synthase 2, also known as COX-2 (cyclo-oxygenase-2)]-positive cells. Overexpression of EP1, as well as EP3, had no effect on LPS-induced Ccl2 mRNA expression. In contrast, overexpression of EP2, as well as EP4, significantly decreased LPS-induced CCL2 expression. These results support the hypothesis that PTGS2-derived prostaglandins, when strongly induced, counter-balance inflammatory processes through the EP2 and EP4 receptors in MCs.
- chemokine CCL2
- cyclo-oxygenase 2 (COX-2)
- lipopolysaccharide (LPS)
- mesangial cell
- prostaglandin EP receptor
- prostaglandin-endoperoxide synthase 2 (PTGS2)
- renal glomerular inflammation
PGE2 (prostaglandin E2) is an important arachidonic acid metabolite generated by sequential action of prostaglandin G/H synthase and prostaglandin E synthase [1–3]. Pro-inflammatory agents such as interleukin-1β  and LPS (lipopolysaccharide) , a cell wall component of Gram-negative bacteria, rapidly induce PGE2 by stimulating PTGS2 [prostaglandin G/H synthase 2, also known as COX-2 (cyclo-oxygenase 2)]. Thus PTGS2 generated PGE2 plays an important role in early inflammatory processes, including diseases associated with inflammation, such as rheumatoid arthritis  and different forms of glomerulonephritis [7–9].
In immune-mediated glomerulonephritis, PGE decreases damage through reduction of glomerular immune complex formation, as well as via reduction of inflammatory cell infiltration [10–11] and deposition of extracellular matrix products .
The physiological effects of PGE2 are mediated through prostaglandin E receptors, called EP receptors. Four EP receptors (EP1 to EP4) are currently known. They are all G-protein-coupled seven-transmembrane receptors, with a strong heterogeneity in respect to the underlying signal transduction pathways. The EP1 receptor is linked to stimulation of intracellular calcium [13,14]. The EP3 receptor exists in multiple splice variants that show different activation patterns; Sugimoto et al.  demonstrated an EP3-mediated inhibition of cAMP through activating a Gi (inhibitory) protein over 15 years ago. Later studies demonstrated that other splice variants are coupled to stimulation of cAMP  or IP3 (myo-inositol 1,4,5-trisphosphate) . The EP2 receptor is coupled to a Gs (stimulatory) protein leading to stimulation of cAMP formation [18,19]. The EP4 receptor is also coupled to a Gs protein, but cAMP formation is weaker than with EP2 activation. In addition, the EP4 receptor is able to activate the phosphoinositide 3-kinase-dependent pathway leading to activation of MAPK (mitogen-activated protein kinase) signalling [20,21]. Thus PGE2 may cause different intracellular responses depending on the specific distribution of the different EP receptors on a particular cell type. In the immune system, through such a cell-specific EP receptor distribution, PGE2 may have different roles with a complex pattern of pro-inflammatory , as well as anti-inflammatory , effects.
CCL2 [chemokine (Cys-Cys motif) ligand 2; also known as MCP-1 (monocyte chemotactic protein-1)], a pro-inflammatory chemokine of the Cys-Cys group, is highly expressed in glomeruli under pathophysiological conditions associated with glomerular infiltration of monocytes/macrophages [24–26]. Prostaglandin G/H synthase metabolites, especially PGEs, reduce glomerular CCL2 expression and thus ameliorate monocyte/macrophage influx in glomerulonephritis [27,28]. MCs (mesangial cells) that constitute a major part of the glomerular structure, and are located between the glomerular capillaries , produce CCL2 in response to diverse inflammatory stimuli [30–32]. Therefore MCs have been identified as early targets of inflammatory processes in a variety of glomerular diseases .
As the role of the different EP receptors in inflammatory renal diseases is unclear, the present study investigates a possible role of EP2 receptor deficiency in LPS-induced nephritis. We focus on the role of EP receptors on CCL2 expression in a simplified experimental model. Our in vitro model system used stimulated MCs, with constitutively enhanced PGE2 secretion and down-regulated EP receptor expression (PTGS2+ cells). By selective overexpression of EP receptors in these cells, their effects on LPS-induced CCL2 expression are examined.
LPS-induced nephritis was induced by intraperitoneal LPS (Escherichia coli O111:B4) injection in WT (wild-type) C57BL/6 (Charles River) and EP2 receptor-knockout mice (EP2ko), a strain which lacks the EP2 receptor  as described previously . The animals were bred under flow-sterile conditions. The experiments were performed with 3 mg of LPS per kg of body weight. Kidneys were examined at 3 and 24 h after LPS injection.
All animal experiments were performed according to national and institutional animal care and ethical guidelines, and have been approved by the local ethical committee.
Cell lines and transient transfections
WT MC lines, as well as an EP2 prostaglandin receptor knockout MC line (EP2koMC), were established from glomeruli of appropriate animals by differential sieving and were characterized as described previously . Both cell lines were treated with 10 μM of the EP4 specific antagonist GW-627368X (Cayman Chemical) with or without 1 μg/ml VC (vector control) plasmid and LPS. PTGS2+ cells were used as described previously . The cell lines were maintained in RPMI 1640, containing 10% (v/v) FBS (fetal bovine serum) and were cultured at 37 °C in a 5% CO2 atmosphere. The medium for PTGS2+ and VC cells was supplemented with 200 μg/ml zeocin or 200 μg/ml G418 respectively (both Invitrogen).
For transient transfection, 2×105 WT or PTGS2+ cells were seeded in six-well plates and cultured overnight. For RNA experiments 2 μg of the appropriate EP expression construct (full-length EP1–EP4 cloned in pcDNA3.1neo; Missouri S&T Resource Center) or 2 μg of the empty pcDNA 3.1neo vector together with 0.5 μg of a β-galactosidase expression construct (Stratagene) were co-transfected for 4 h by using the Plus™/Lipofectamine™ LTX system (Invitrogen) according to the manufacturer's instructions. To recover cells, and to ensure expression of the transfected constructs, transfection medium was removed and cells were cultured overnight in RPMI 1640 containing 0.5% FBS. Then, LPS stimulation was performed in serum-free RPMI 1640 medium.
Microdissection of glomeruli
For microdissection of glomeruli, 12 μm thick cryosections were generated and stained with Cresyl Violet. Approx. 150 glomeruli sections were cut directly out of the tissue by using a Palm Microm Beam microscope (Palm Microlaser Technologies) and captured in tissue lysis buffer (RLT; Qiagen).
RNA isolation, cDNA synthesis and qRT-PCR (quantitative real-time PCR) analysis
Isolation of total RNA from MCs was performed using the Nucleo Spin RNA-II Kit (Macherey & Nagel) and isolation of total RNA from microdissected glomeruli was performed using the Micro RNA Preparation Kit (Quiagen), both according to the manufacturer's instructions.
Total RNA from 150 microdissected glomeruli or 200 ng of MCs was reverse transcribed with 100 units of MMLV (Moloney-murine-leukaemia virus) reverse transcriptase (Invitrogen) for 50 min at 37 °C. Enzyme activity was subsequently stopped by incubation for 15 min at 70 °C. cDNA corresponding to 150 glomeruli or 10 ng of total cellular RNA respectively was subjected to SYBR Green-based qRT-PCR using the SYBR Green JumpStart kit (Sigma–Aldrich). To assess differences in cDNA quality, the probes were normalized to 18S rRNA expression. For CCL2 the forward primer was 5′-CTCAGCCAGATGCAGTTAATG-3′ and the reverse primer was 5′-TTCTCCAGCCGACTCATTGG-3′. For 18S rRNA the forward primer was 5′-CACGGCCGGTACAGTGAAAC-3′ and the reverse primer was 5′-AGAGGAGCGAGCGACCAAA-3′.
Quiescent VC and PTGS2+ cells were treated with 6–48 nmol of 3H-labelled PGE2 (Amersham Biosciences) in the presence of 500 nmol of cold PGE2 (Cayman Chemical) for 30 min. Thereafter, cells were washed four times in PBS and then trypsinized. The 3H content of the cell solution was subsequently measured in a β-scintillation counter (Packard).
Conditioned cell medium was collected and the level of CCL2 was quantified using the Instant ELISA kit (Bender Medsystems) according to the manufacturer's instructions. The CCL2 level was normalized to total cellular protein of the appropriate sample.
Western blot analysis
Western blot analysis was performed as described previously . All anti-(human EP) primary antibodies were obtained from Cayman Chemical and the anti-(human β-actin) was purchased from Sigma–Aldrich. All the antibodies were used at a dilution of 1:1000. For detection, an anti-rabbit-IgG antibody (for all of the EP receptors) or an anti-mouse-IgG (for β-actin) conjugated to alkaline phosphatase (Southern Biotechnology) at a concentration of 1:5000 was used.
The data are presentated as the means±S.D. Statistical significance between different groups was first tested with the non-parametric Kruskal–Wallis test. Individual groups were subsequently tested using the Wilcoxon–Mann–Whitney test. A P-value<0.05 was considered significant.
EP2 receptor deficiency modulates Ccl2 expression in LPS-induced nephritis
To examine whether EP2 receptor deficiency influences the expression of the pro-inflammatory chemokine CCL2 in LPS-induced nephritis, nephritic WT and EP2ko mice were killed 3 and 24 h after LPS injection. Total RNA was isolated from the renal cortex and from microdissected glomeruli and qRT-PCR was performed to determine the level of Ccl2 mRNA expression.
LPS strongly induced Ccl2 mRNA expression in the renal cortex of WT and EP2ko mice and there was no difference between WT and EP2ko mice at both time points in the crude cortex preparations (results not shown). In microdissected glomeruli, Ccl2 mRNA expression was up-regulated 3 h after LPS injection in EP2ko mice compared with WT mice (2.53±1.14-fold) (Figure 1A), but by 24 h post-LPS injection the expression levels were similar in both WT and EP2ko mice (Figure 1A).
LPS-induced CCL2 expression is modified by different EP receptors in MCs
To confirm the participation of EP receptors in the modulation of LPS-stimulated Ccl2 mRNA expression on a cellular level, quiescent WT mouse MCs and EP2ko MCs were stimulated with 1 μg/ml LPS for 3 h in the presence (10 μM) or absence of an EP4 antagonist GW-627368X, after pre-treatment for 1 h with GW-627368X.
Figure 1(B) shows that LPS stimulates Ccl2 mRNA expression in WT cells (127.65±17.32-fold). In EP2ko cells, lacking the EP2 receptor, the LPS-induced stimulation of Ccl2 mRNA expression is enhanced (299.54±150.96-fold). This LPS-induced stimulation of Ccl2 is comparable with that in WT in the presence of 10 μM EP4 antagonist (310.6±55.3-fold), but is even more pronounced in EP2ko cells in the presence of the EP4 antagonist (786.25±311.3-fold). Treatment of both cell lines with the EP4 antagonist in the absence of LPS did not show any significant effect (results not shown).
Figure 1(C) demonstrates CCL2 protein secretion into cell culture medium over 6 h in the same experimental setting. LPS-induced WT cells secrete 31.8±5.2 pg of CCL2 per mg of total protein, whereas EP2ko cells secrete 199.7±28.3 pg of CCL2 per mg of total cellular protein. In the presence of the EP4 antagonist CCL2 protein secretion is further enhanced in WT (217.2±26.5 pg/mg) and in EP2ko cells (290.7±40.1 pg/mg). Incubation of the cell lines with GW-627368X alone showed no significant effects (results not shown).
Transient overexpression of EP2 and EP4 receptors influences LPS-induced CCL2 expression in WT cells
The data depicted in Figure 1(A)–1(C) suggest that EP2 and EP4 receptors modulate LPS-induced Ccl2 gene expression. To further test this hypothesis, EP2 and EP4 receptors were overexpressed in WT cells. WT cells were transiently transfected with either plasmids expressing the EP2 or EP4 receptor, using mock-transfected cells as a control. After 1 day these cells were LPS-stimulated for 3 h, or were pre-incubated for 1 h with 1 μM PGE2, and were then LPS-stimulated for another 3 h in the presence of PGE2. As shown in Figure 1(D), overexpression of EP2 (0.74±0.1-fold) or EP4 receptors (0.83±0.1-fold) in WT cells slightly reduced the level of LPS-induced Ccl2 mRNA expression. In the presence of 1 μM PGE2 the LPS-induced reduction of Ccl2 mRNA is further reduced (EP2, 0.54±0.14-fold; EP4, 0.5±0.17-fold).
Cells with constitutive PGE2 secretion show a reduced cell-surface expression of EP receptor due to EP receptor down-regulation
The results of the transient overexpression of EP2 and EP4 receptors in WT cells (Figure 1D) indicate that there is a possible role of PGE2-stimulated EP receptors in the regulation of Ccl2 gene expression. To examine this effect in more detail, we chose to use a cell line with constitutive PGE2 secretion (PTGS2+)  in order to overcome the problem of using WT cells which have unstable exogenous delivery of PGE2 (due to lability of dissolved PGE2 stocks) and therefore possible ongoing regulation of EP receptors. We expect that the permanent availability of freshly generated, bioactive PGE2 in PTGS2+ cells is an important prerequisite for examining overexpression of different EP receptor pathways to enable understanding of an individual EP receptor function.
To detect possible EP receptor down-regulation in the presence of enhanced PGE2, 3H-labelled PGE2 binding was examined in PTGS2+ cells and VC cells. These cells were pulsed for 30 min with 6–48 nmol of 3H-labelled PGE2 in serum-free culture media. Total PGE2 binding was significantly reduced in PTGS2+ cells concentrations of PGE2 higher than 24 nmol (Figure 2A). This suggests that PGE2 significantly reduces EP receptor cell-surface expression.
In order to confirm the binding results on EP receptor levels, appropriate mRNA expression and immunoblot analysis was performed. As shown in Figures 2(B) and 2(C), the expression of EP1, EP2 and EP4 receptors were significantly down-regulated in PTGS2+ cells at the mRNA and protein level. EP3 is also down-regulated at the mRNA level, but the available antibody did not detect any isoform in MCs and we were therefore unable to directly assess protein level.
In accordance with the results on enhanced CCL2 protein secretion during EP2 or EP4 knockdown (Figure 1), CCL2 protein secretion from LPS-treated PTGS2+ cells (674.6±184.8 pg of CCL2 per mg of total protein) is significantly increased compared with VC cells (68.8±14.8 pg of CCL2 per mg of total protein) (Figure 2D).
Overexpression of EP receptors and CCL2 expression
In order to examine the exact role of each EP receptor, overexpression experiments in the PTGS2+ cell line were performed. EP1, EP2, EP3 or EP4 was transiently overexpressed in PTGS2+ cells and expression of the different EP receptors was examined 24 h post-transfection. As shown in Figure 3, mRNA expression is enhanced for all of the receptors, as well as protein expression of the EP1, EP2 and EP4 receptor (protein expression of the EP3 receptor could not be detected with the available, antibody).
At 24 h post-transfection PTGS2+ cells were further treated with 1 μg/ml LPS for 3 h and compared with LPS-treated mock-transfected PTGS2+ cells. As depicted in Figure 4, overexpression of EP1 (mRNA levels increase 1.22±0.43-fold) as well as EP3 (mRNA levels increase 1.42±0.53-fold) hardly influenced LPS-induced CCL2 protein secretion (647.6 ±110.6 pg/mg and 669.4±170.2 pg/mg respectively). In contrast, overexpression of EP2 (mRNA levels decrease 0.22±0.05-fold), as well as EP4 receptors (mRNA levels decrease 0.35±0.07-fold) caused a significant decrease in LPS-induced CCL2 protein secretion (protein: 185.7±23.9 pg of CCL2 per mg of total protein and 197.2±28.5 pg of CCL2 per mg of total protein respectively).
A characteristic feature of many diseases associated with inflammatory processes, and also different forms of glomerulonephritis, is the early up-regulation of PTGS2-dependent PGE2 formation [6–9]. This up-regulation can be mediated, at least partly and experimentally, by the action of pro-inflammatory inducers such as interleukin-1β  or LPS . However, PGEs produced during the inflammatory process can also down-regulate expression of pro-inflammatory chemokines, as has been demonstrated for CCL2 in experimental glomerulonephritis [27,28]. A complex pattern of pro-inflammatory , as well as anti-inflammatory , effects may therefore be obtained in response to PGE2. PGE2 may influence inflammatory processes differently, depending on the model and type of disease. However, the role of PGE2 is also dependent on the time-point of its expression during the course of the inflammation and the distribution of receptors on the surfaces of the cells.
mRNA of the EP1 and EP4 receptor, and their possible roles under high-glucose conditions, was first reported in rat MCs in culture by Ishibashi et al. . Other authors subsequently described EP2 receptor mRNA in rat MCs in culture and conflicting results were published on the presence of EP3 receptor mRNA in MCs [38,39]. To date these reports do not give a clear picture of the EP receptor distribution, or their possible roles, in MCs. The role of EP receptors in experimental or human nephritis is even more unclear. Experiments using a model of nephrotoxic nephritis suggest an enhanced susceptibility of renal dysfunction in mice lacking the EP1 receptor . Evidence for a protective role of the EP4 receptor in nephritis comes from a study using an EP4 receptor agonist in an anti-glomerular basement membrane model of nephritis in mice  and a study using an EP4 receptor antagonist in the acute puromycin aminonucleoside model of nephritis in rats .
We were interested as to whether mice deficient in EP2 receptor expression show a different inflammatory response compared with WT mice. We chose the LPS-induced nephritis model for our examinations, because LPS is a rapid inducer of the inflammatory process, strongly invoking secretion of CCL2. The pro-inflammatory chemokine CCL2 is known to be expressed in the early phases of different forms of immune-mediated glomerulonephritis [24–26] and its expression may be modulated through PGE2 [27,28]. The experiments in the present study demonstrate that glomerular CCL2 expression is enhanced in the early phase of LPS-induced nephritis in EP2-deficient mice compared with WT mice. This result suggests that in this model there is a role for PGE2 and EP receptors in CCL2 regulation. Furthermore, the observed effect of EP2 deficiency is restricted to glomeruli, implying that the potential role is specific for MCs.
A previous study with MCs suggested that of one effect of PGE in the regulation of inflammation was that it mediated cAMP stimulation . Thus, in addition to the EP2 receptor, the EP4 receptor is a good candidate for mediating similar effects. Unfortunately, standard constitutive EP4-knockout mice die prenatally [43–45] and conditional EP4-knockout MCs have not been generated. Therefore, in order to gain further insight into the role of EP receptors in the PGE2-mediated inflammatory processes, we further examined LPS-induced Ccl2 gene expression in cultured MCs where specific EP receptors were overexpressed.
Initial cell culture experiments were performed in MCs lacking the EP2 receptor (EP2ko) and results compared with WT MCs. In the EP2ko cells Ccl2 gene expression is enhanced in response to LPS. Corresponding results were obtained in LPS-induced WT cells in the presence of a specific EP4 receptor antagonist. In addition, blocking EP4 receptor in EP2ko cells further enhanced LPS-induced Ccl2 gene expression. Thus the availability of active EP2 and EP4 receptors seems to be necessary to induce the strong inflammatory effects of LPS. In order to test this hypothesis, we performed transient overexpression of each EP receptor and measured Ccl2 gene expression.
Prostaglandins, including PGE2, are highly oxygen-sensitive fatty acids (G. Zahner, unpublished work). This means that when overexpressing EP receptors, significant amounts of bioactive extracellular PGE2 are required. We therefore decided to test the effect of EP receptor overexpression in a MC line that constitutively generates bioactive PGE2, due to the constitutive overexpression of PTGS2 (PTGS2+ cells) . PGE2-binding assays demonstrated that there was significantly reduced binding of PGE2 to the cell surface in PTGS2+ cells, and that the expected down-regulation of EP receptors, due to the high extracellular PGE2 content, occurred. Stimulation with LPS strongly induced Ccl2 gene expression in these cells. Thus these cells have a low basal EP receptor cell-surface expression and have high amounts of bioactive PGE2 (thereby eliminating the need for further stimulation with labile exogenous PGE2). The PTGS+ cells therefore make a good in vitro model in which to study the effects of individual EP receptors on CCL2, through transient overexpression of each receptor. Overexpression of the EP2 (which signals through activation of the second messenger cAMP) and the EP4 receptor (signalling through phosphoinositide 3-kinase) results in a significant reduction of LPS-induced Ccl2 mRNA expression. This observation confirms the initial EP2 receptor-knockdown and EP4 receptor-blocking experiments. In contrast, neither EP3 nor EP1 receptor overexpression affects Ccl2 mRNA expression. The EP2 receptor-dependent reduction of Ccl2 mRNA expression is more potent than the appropriate reduction observed with EP4 receptor overexpression. Differences between the EP2 receptor activation of cAMP downsteam pathways [18,19], mainly activation of protein kinase A , and the possible EP4 receptor-dependent activation of IP3 kinase accompanied with a subsequent activation of MAPKs  may explain these differences. Examinations of these different relevant signalling pathways will help to clarify the discrepancies between the two receptors.
In conclusion, in vivo studies of LPS-induced nephritis suggest that PGE2, and its specific E2 receptor, is involved in the acute phase of inflammation. By using an in vitro model with down-regulated EP receptors, and therefore strongly reduced background noise, the EP2 and EP4 receptors, but not the EP1 and EP3 receptors, were identified as important modulators of PGE2-mediated inflammatory effects, as measured by Ccl2 mRNA expression and protein secretion. The down-regulation of EP2 and EP4 receptors results in an imbalance in the inflammatory state of MCs. The results of the present study identify the EP2 and EP4 receptors as potential feedback mediators of inflammatory PGE2 effects in these cells. EP2 and EP4 receptors may be important mediators of the PTGS2 metabolite-dependent down-regulation of the expression of pro-inflammatory chemokines, such as CCL2, in glomerulonephritis.
Gunther Zahner planned the study, performed experiments, analysed the results and wrote the manuscript. Melanie Schaper performed experiments. Ulf Panzer analysed the results. Malte Kluger and Friedrich Thaiss provided the expertise on the animal disease model of LPS-induced nephritis. Rolf Stahl provided funding and laboratory space. André Schneider planned the study, analysed the results and wrote the manuscript.
This work was supported by departmental funding to Zentrum Innere Medizin.
Abbreviations: CCL2, chemokine (Cys-Cys motif) ligand 2; EP, prostaglandin E receptor; EP2ko, prostaglandin E receptor 2 knockout; FBS, fetal bovine serum; IP3, myo-inositol 1,4,5-trisphosphate; LPS, lipopolysaccharide; MAPK, mitogen-activated protein kinase; MC, mesangial cell; PGE, prostaglandin E; PTGS2, prostaglandin-endoperoxide synthase 2; qRT-PCR, quantitative real-time PCR; VC, vector control; WT, wild-type
- © The Authors Journal compilation © 2009 Biochemical Society