Abstract
Chemokines and their receptors play an important role in the pathogenesis of acute and chronic glomerular inflammation. However, their expression pattern and function in glomerular podocytes, the primary target cells in a variety of glomerulopathies, have not been investigated as of yet. Using RT-PCR, we now demonstrate the expression of CCR4, CCR8, CCR9, CCR10, CXCR1, CXCR3, CXCR4, and CXCR5 in cultured human podocytes. Stimulation of these receptors induced a concentration-dependent biphasic increase of the free cytosolic calcium concentration in podocytes in culture. In addition, we demonstrate that podocytes release IL-8 in the presence of FCS and that IL-8 down-regulates cell surface CXCR1. Chemokine stimulation of the detected CCRs and CXCRs increased activity of NADPH-oxidase, the primary source of superoxide anions in podocytes. Immunohistochemistry studies revealed only diffuse and weak CXCR expression in healthy human glomerula. In contrast, in membranous nephropathy, a characteristic podocyte disorder, the expression of CXCR1, CXCR3, and CXCR5 is up-regulated in podocytes. In conclusion, podocytes in culture and podocytes in human kidney sections express a set of chemokine receptors. The release of oxygen radicals that accompanies the activation of CCRs and CXCRs may contribute to podocyte injury and the development of proteinuria during membranous nephropathy.
The podocyte is a highly specialized cell which constitutes a crucial component of the glomerular filtration barrier. Podocyte damage leads to the retraction of their foot processes, resulting in proteinuria. Especially in diabetic nephropathy, minimal change nephropathy, membranous nephropathy (MGN) ,4 and focal segmental glomerulosclerosis, podocytes are the primary target of injury (1). The precise mechanisms that lead to podocyte damage and proteinuria in glomerular diseases are only roughly understood. It has been suggested that presently unknown circulating mediators might affect podocyte functions and cause the retraction of foot processes and, thus, proteinuria in minimal change nephropathy and focal segmental glomerulosclerosis (2). In those glomerular diseases that involve podocyte injury, it has been suggested that cytokines mediate the inflammatory processes that ultimately result in proteinuria. For example, in Heymann nephritis, an experimental rat model for MGN, depletion of CD8 cytotoxic T cells prevents proteinuria, indicating that cytokines secreted by CD8 cytotoxic cells may be involved in podocyte injury (3). During puromycin aminonucleoside nephrosis, an experimental model for minimal change nephropathy, an increase of IFN-inducible protein (IP-10), monocyte chemoattractant protein (MCP) 1, MCP-3, and T cell activation gene 3 mRNA expression has been reported (4, 5, 6).
Chemokines are a group of small peptides that are subdivided into four families, comprising >50 ligands with at least 17 different receptors. These chemokine families are defined by the presence of either a C, a CC, a CXC, or a C′C residue at the amino terminus of the protein. The largest of these subfamilies is the CXC chemokines, in which two amino-terminal cysteines are separated by a nonconserved amino acid, and the CC chemokines, in which two amino-terminal cysteines are juxtaposed (7). Within the glomerulum, chemokines and their receptors are expressed in infiltrating cells as well as in resident glomerular cells. Glomerular-produced chemokines seem not only to induce recruitment of inflammatory cells, but can also alter functions of resident glomerular cells, such as the formation of extracellular matrix (8). Therefore, chemokines may play an important role in the events leading to podocyte injury and proteinuria. In this study, we investigated the expression and function of chemokine receptors in cultured podocytes and in human kidney sections.
Materials and Methods
Cell culture
Human podocytes were isolated from normal human kidney sections obtained from renal carcinoma patients undergoing nephrectomy. Primary cultures of podocytes were established as follows: Intact tissue was passed through steel sieves with decreasing pore sizes of 450 and 180 μm. Glomerula were collected using a third sieve with a pore size of 120 μm and counted under a light microscope (×400). Encapsulated glomerula were absent upon visual inspection. Glomerula were suspended in DMEM containing 10% heat-inactivated FCS, 2.5 mM glutamine , 0.1 mM sodium pyruvate, 5 mM HEPES buffer, 1 mg/ml streptomycin, 100 U/ml penicillin, 0.1× nonessential amino acids (100×; all Seromed, Berlin, Germany), insulin, transferrin, and a 5 mM sodium selenite supplement and then plated at a concentration of 100 glomerula/cm2 onto collagen-coated petri dishes (Greiner, Nürtingen, Germany). Glomerula were incubated at 37°C and 5% CO2 in air. After 4 days, cell colonies began to sprout around the glomerula. Cell colonies were excised and incubated in a tube containing 5 ml 0.2% collagenase IV (Sigma-Aldrich, Deisenhofen, Germany) at 37°C for 30 min and were then washed and plated in 25-cm2 culture flasks.
Cells showed an epithelial morphology with a polyhedral shape when confluency was reached and were immunohistologically characterized as podocytes. They stained positive for Wilm’s tumor Ag (WT1) and expressed nephrin, markers which are only found in podocytes in the adult kidney. Moreover, cells were negative for factor 8 related Ag, a marker for endothelial cells. Cells between passages 15 and 23 were seeded at 37°C onto collagen A (Biochrom, Berlin, Germany)-coated plates and cultured in standard RPMI 1640 medium containing 10% FCS, 100 U/ml penicillin, and 1 mg/ml streptomycin. In another set of experiments, early passages (passage 1 and passage 4) of cultured podocytes were used to confirm the functional expression of the chemokine receptors in short-term cultured cells. In addition, a conditionally immortalized human podocyte cell line demonstrating nephrin and podocin expression was used (9)
Cultures of primary human podocytes were infected with retrovirus-containing supernatants from the packaging cell line (PA317). The retroviral construct consisted of a SV40 large T-Ag gene containing both the tsA58 and the U19 mutations (10). Infection, selection, and continuous culture were conducted at 33°C. A single-cell clone was used for all of the experiments described.
Induction of differentiation
Subsequently, cells were grown on type I collagen-coated flasks layered with glass coverslips for the purpose of immunostaining. Cells were then plated onto the flasks and grown either at the “permissive” temperature of 33°C (in 5% CO2) to promote cell propagation as a cobblestoned phenotype or at the “nonpermissive” temperature of 37°C (in 5% CO2) to inactivate the SV40 T-Ag and allow the cells to differentiate.
As a positive control for CXCR2 expression, we used human lung microvascular endothelial cells in one series of experiments (11). Cells were cultured as previously described (11).
Expression of CCR and CXCR mRNA in cultured human podocytes and human glomerula
The RNA preparation, the reverse transcription and the PCR amplification were performed according to the method described recently (12). In brief, the total RNA from cultured human podocytes or human glomerula was isolated with guanidinium/acid phenol/chloroform extraction and the amount of RNA was measured with spectrophotometry. For first-strand synthesis, 2 μg of total RNA was mixed in 5× reverse transcription buffer and incubated with 5 U of DNase I for 15 min. After termination, the reaction was completed with 0.5 mM dNTP, 0.5 μM sequence-specific primers, 10 mM DTT, and 200 U of superscript II transcriptase (reverse transcriptase was omitted in some experiments to control for the amplification of contaminating DNA). The reverse transcription was performed at 42°C for 1 h followed by a denaturation at 95°C for 5 min. cDNA was purified from amplification reaction and solved in 30 μl of 10 mM Tris-buffer (pH 8). PCR was performed in duplicates in a total volume of 20 μl, each containing 2 μl of reverse transcription reaction and 18 μl of PCR master mixture containing 10 pmol each of sense and antisense primer and 2.5 U of Taq DNA polymerase. The cycle profile included denaturation for 30 s at 94°C, annealing for 30 s at 60°C, and extension for 30 s at 72°C. In the experiments for the analysis of the mRNA expression of CXCR1, 36 cycles of PCR were performed. In all other experiments, 30 cycles of PCR were performed.
The amplification products of 10 μl from each PCR were separated on a 1.5% agarose gel, stained with ethidium bromide, and visualized by UV irradiation. PCR amplification of reverse transcription reactions without reverse transcriptase revealed no PCR product, thereby excluding amplification of genomic DNA. The identity of amplification products was determined by dideoxy sequencing. The primers used are shown in Table I⇓.
PCR primers of CCR and CXCR chemokine receptors and the specific podocyte proteins WT1 and nephrin
Semiquantitative analysis
The amplification products of 10 μl of PCR product were separated on a 1.5% agarose gel, stained with ethidium bromide, visualized with UV irradiation, and photographed with Polaroid film 667. The photographs were taken to evaluate the band densities by volume integration using a Hewlett Packard IIcx flatbad scanner and computer based imaging software (Image Quant; Molecular Dynamics, Krefeld, Germany). The data were normalized on the GAPDH mRNA expression.
Measurements of the free cytosolic intracellular Ca2+ concentration ([Ca2+]i)
Measurements of cytosolic calcium with the Ca2+-sensitive dye fura 2 (Sigma-Aldrich) were performed in podocytes on an inverted fluorescence microscope setup (13). The system allows fluorescence measurements at the single-cell level at three excitation wavelengths. The field of measurement can be set between a diameter of 2 and 300 μm with an adjustable pinhole. A time resolution of up to 200 Hz was achieved by using a high-speed filter wheel and a single photon counting tube (Hamamatsu H63460-04; Hamamatsu, Herrsching, Germany). The autofluorescence signal of cells that had not been loaded with fura 2 was measured and subtracted from the results obtained in fura-2-loaded cells. This had no effect on the bandwidth of the measurements. A calibration of the fura-2 fluorescence signal was attempted at the end of each experiment by using Ca2+ ionophore ionomycin (1 μM) and low and high Ca2+ buffers. [Ca2+]i was calculated from the fluorescence ratio 340:380 nm according to the equation described by Grynkiewicz et al. (14).
Measurement of IL-8 release
15) and used to normalize the IL-8 release on the protein content per well.
Measurement of NADPH-oxidase activity
Podocytes were rinsed twice with ice-cold PBS, scraped from the wells, and resuspended in 2 ml of Krebs buffer. After centrifugation at 750 × g for 5 min at 4°C, the pellet was resuspended in 1.5 ml of fresh Krebs buffer (pH 7.35; containing 99 mM NaCl, 4.7 mM KCl, 1.8 mM CaCl2, 1.2 mM MgCl2, 25 mM NaHCO3, 1.03 mM K2HPO4, 20 mM sodium-HEPES, and 11.1 mM glucose). Cells were centrifuged as above and then resuspended in 0.6 ml of Krebs buffer. The luminescence buffer contained 5 μM lucigenin and 0.1 mM NADPH as the substrate.
To calculate the amount of superoxide produced, total counts were generated by integrating the area under the signal curve. These values were compared with a standard curve which was generated by using xanthine/xanthine oxidase as described elsewhere (12).
Superoxide generation was expressed as nanomoles of O2− generated per milligram of cellular protein per minute as described earlier (16). Protein content of the cell suspension was measured with the Lowry method (15).
Immunhistochemistry of cultured human podocytes
The immunolabeling was done as previously described (17). Briefly, coverslips were fixed with 2% paraformaldehyde and 4% sucrose in PBS, for 10 min, and were then permeabilized with 0.3% Triton X-100 (Sigma-Aldrich) in PBS for 10 min. Nonspecific binding sites were blocked with 4% FCS plus 0.1% Tween 20 (Sigma-Aldrich) in PBS for 30 min. Primary and secondary Abs were applied in the appropriate dilutions according to standard techniques, and the coverslips were mounted on glass slides with 15% Mowiol (Calbiochem, La Jolla, CA) and 50% glycerol in PBS. Further washes and mounting were conducted as above. Images were obtained using a Leica photomicroscope attached to a Spot 2 slider digital camera (Diagnostic Instruments, Solms, Germany) and processed with Adobe PhotoShop 5.0 software (Adobe Systems, Unterschleissheim, Germany).
Immunohistochemistry of kidney cortex
Normal kidney cortex and cortex from two patients with membranous nephropathy were obtained from nephrectomy specimens and snap frozen in liquid N2
Chemicals
Statistics
The data are given as mean values ± SEM, where n refers to the number of experiments. The average of the effect of the agonist before and after the experiment was taken as control. A paired t test was used to compare mean values within each experimental series. A p ≤ 0.05 was accepted to indicate statistical significance.
Results
Human cultured podocytes and human glomerula express mRNA for CCR and CXCR subtypes
By using RT-PCR, we studied the expression of CCR and CXCR subtypes in human cultured podocytes and human glomerula with sequence-specific primers for each different receptor subtype (Table I⇑). Fig. 1⇓A shows ethidium-bromide-stained agarose gel electrophoreses of PCR products for CCR4, CCR8, CCR9, and CCR10 in cultured human podocytes. Fig. 1⇓B shows ethidium bromide-stained agarose gel electrophoreses of PCR products for CXCR1, CXCR3, CXCR4, and CXCR5 in cultured human podocytes. In contrast to all other CCRs and CXCRs, which were detectable with 30 cycles of PCR, CXCR1 could only be constantly detected with 36 cycles of PCR.
Expression of mRNA for CCRs (A) and CXCRs (B) in human cultured podocytes and human glomerula. Fragments of the different chemokine receptors were amplified using specific primers (see Table I⇑) and subjected to agarose gel electrophoresis. Lanes on the left side indicate base pair length. Note that podocyte bands with the expected sizes are present in the CCR4, 8, 9, and 10 lane and in the CXCR1, CXCR3, CXCR4, and CXCR5 lane, whereas in human glomerula (lower panels), bands of CCR1–10 and CXCR1–5 are positive. In both human cultured podocytes and human glomerula, the bands for WT1 and nephrin, which are only expressed in podocytes, were present. Sequence analysis of the resulting amplification products revealed the sequence identity of the fragments.
Sequence analysis of the resulting amplification products revealed the sequence identity of the fragments. mRNA could not be detected for CCR1–3 and CCR5–7 and CXCR2 in human podocytes using 30 or 36 cycles of PCR.
It is known that proinflammatory cytokines such as IL-1β can stimulate the expression of chemokine receptors in other cell types (18). IL-1 β has been shown to activate the transcription factor NF-κB in podocytes (19). In addition, IL-1β expression is up-regulated in podocytes in Heymann nephritis. To test whether IL-1β might induce mRNA expression for those CCRs and CXCRs which could not be found by RT-PCR, podocytes were treated with Il-1β (10 ng/ml) for 6 and 18 h, and mRNA expression of CCRs and CXCRs was studied. IL-1β did not induce the expression of either CCR1–3 or CCR5–7 nor did it induce the expression of CXCR2 (n = 3, data not shown).
In human glomerula, the PCR products of CCR1–CCR10 (Fig. 1⇑A, lower panel) and CXCR1–CXCR5 (Fig. 1⇑B, lower panel) could be detected. In addition, we studied the mRNA expression of CCRs and CXCRs in conditionally immortalized, differentiated human podocytes, which may better mimic podocytes in vivo. These podocytes showed the same expression pattern of CCRs and CXCRs, namely, the CCR4, CCR8–10, and CXCR1, CXCR3, CXCR4, and CXCR5 (data not shown).
CCR ligands increase the free cytosolic calcium concentration [Ca2+]i [infi] in human cultured podocytes
Addition of the CCR4 agonist macrophage-derived chemokine (MDC, n = 33), the CCR8 agonist I-309 (n = 33), the CCR9 agonists TECK (n = 6) and HCC-1 (n = 2), and the specific CCR10 agonist human CCL28 (n = 4) (20, 21) to single podocytes loaded with fura-2 resulted in a reversible increase of [Ca2+]i. The left panel of Fig. 2⇓ shows the typical original fluorescence measurements obtained from single podocytes exposed to MDC (A, 10 nM), I-309 (B, 100 nM), HCC-1 and TECK (C, 2.5 μg/ml and 100 nM, respectively), and CCL28 (D, 2 μg/ml). The right panel of Fig. 2⇓ gives the concentration response curves (A, B, and D) or the summary of the experiments (C and D) for the different CCR agonists.
Fluorescence recordings showing the effect of CC chemokines on [Ca2+]i in podocytes. Left panel, The original recordings of the CC chemokines induced an increase of [Ca2+]i in human podocytes: CCR4 agonist MDC (A), CCR8 agonist I-309 (B), CCR9 agonists HCC-1 and TECK (C), and CCR10 agonist CCL28 (D). Right panel, Concentration response curves or summary of the effects of the CCR ligands. Numbers of observations are shown above the curve. ∗, Statistical significance.
CXCR ligands increase the free cytosolic calcium concentration [Ca2+]i in human cultured podocytes
The CXCR3 agonist MIG (n = 32), the CXCR4 agonist SDF-1α (n = 32), and the CXCR5 agonist BCA-1(n = 46) induced a reversible increase of [Ca2+]i in podocytes. In contrast, different concentrations of IL-8 had no effect on [Ca2+]i (370 μM, n = 5; 37 μM, n = 3; 3.7 μM, n = 2) in podocytes grown in the presence of FCS. After the addition of IL-8, SDF-1α as a positive control induced an increase of [Ca2+]i in podocytes (Fig. 3⇓A, left panel). The [Ca2+]i response to MIG, SDF-1α, and BCA-1 was concentration dependent with an EC50 value of 85 nM for MIG, 0.6 nM for SDF-1α, and 10 nM for BCA-1, respectively. The left panel of Fig. 3⇓ shows typical original fluorescence measurements obtained from single podocytes exposed to IL-8 (A, 370 μM), MIG (B, 100 nM), SDF-1α (C, 10 nM), or BCA-1 (D, 100 nM). The right panel of Fig. 3⇓ gives the concentration response curves for the different CXCR agonists.
Fluorescence recordings showing the effect of CXC chemokines on [Ca2+]i in podocytes. A, IL-8, an agonist of CXCR1 and CXCR2, respectively, had no effect on [Ca2+]i of human podocytes grown in FCS. After addition of IL-8, SDF-1α as a positive control induced an increase of [Ca2+]i. B–D, left panel, MIG, SDF-1α, and BCA-1, ligands for CXCR3, CXCR4, and CXCR5, respectively, induced an increase of [Ca2+]i in human podocytes. Right panel, Concentration response curves of the effects of the CXCR ligands. Numbers of observations are shown above the curve. ∗, Statistical significance.
To test whether early passages of human podocytes express the same set of functional chemokine receptors, the measurement of [Ca2+]i was performed on these cells and revealed the same pattern of calcium responses to CXCR ligands (passages 1 and 4: IL-8 (3.7 μM, n = 10; increase of [Ca2+]i: 0 ± nM); MIG (0, 5 μM, n = 5; increase of [Ca2+]i: 136 ± 30 nM), SDF-1α (10 nM, n = 6; increase of [Ca2+]i: 155 ± 41 nM), and BCA-1 (100 nM, n = 8; increase of [Ca2+]i: 173 ± 36 nM); data not shown).
Immunolabeling of CXCR1–5 in human cultured podocytes
Immunolabeling was performed with human podocytes and conditionally immortalized human podocytes cultured in 5% FCS. Both cultured human podocytes and immortalized, differentiated podocytes revealed an identical pattern of CXCR staining:
Fig. 4⇓ shows the CXCR staining of the human immortalized podocyte cell line. We detected a positive staining for CXCR3–5 (D, F, and G, n = 4), whereas no staining was seen for CXCR1 (B, n = 4) and CXCR2 (C, n = 4). As a negative control, we used mouse IgG as the primary Ab in these experiments (Fig. 4⇓A).
Immunofluorescence studies on the expression of CXCR1–5 in immortalized, differentiated human podocytes. Positive staining for CXCR3–5 (D, F, and G). In podocytes cultured in FCS, no signal for the CXCR1 and CXCR2 (B and C) was detectable. Mouse IgG as negative control for these experiments (A).
Effect of FCS on IL-8 release and expression of CXCR1 in podocytes
To prove whether podocytes themselves may release chemokines, the expression of mRNA for IL-8, MIG, IP-10, SDF-1α, and BCA-1 was investigated using RT-PCR. Fig. 5⇓A shows that podocytes grown in the presence of 10% FCS express mRNA for IL-8 but not for MIG, IP-10, SDF-1α, or BCA-1. Fig. 5⇓B shows that IL-8 protein is released from podocytes after incubation with different concentrations of FCS (1 and 10%) and that only small amounts are released in 0% FCS (n = 6). FCS alone did not contain any detectable IL-8 (data not shown). To further investigate whether CXCR1 or CXCR2 may have been down-regulated by the release of IL-8 in podocytes that have been grown in the presence of FCS, the protein expression of CXCR1 and CXCR2 was investigated in FCS-starved podocytes. Indeed, immunofluorescence staining of starved podocytes showed an expression of CXCR1 but not CXCR2 in immortalized, differentiated podocytes (Fig. 5⇓C) or in human podocytes (data not shown). For the positive control of CXCR2 staining, we used human lung microvascular endothelial cells (Fig. 5⇓C). In FCS-starved podocytes, IL-8 caused a concentration-dependent increase of [Ca2+]i (Fig. 5⇓D).
Effect of FCS on IL-8 release and expression of CXCR1 in podocytes. A, Expression of mRNA for IL-8 but not MIG, IP-10, SDF-1α, or BCA-1 was detected in human podocytes. B, Detection of the release of IL-8 protein from podocytes after incubation with different concentrations of FCS (1% and 10%). C, Immunofluorescence staining showed expression of CXCR1 but not CXCR2 in human podocytes. Mouse IgG was used as a negative control for CXCR1 staining and human microvascular endothelial cells as positive control for CXCR2 staining. D, Concentration response curve of the effect of IL-8 on [Ca2+]i in podocytes that were grown without FCS for 24 h. Numbers of observations are shown above the curve. ∗, Statistical significance. E, IL-8 induced an increase of [Ca2+]i in the presence of serum-free medium (0% FCS) and in the presence of FCS-containing medium (1% FCS) with the IL-8-neutralizing Ab (n-IL-8 mab). F, Fragments of CXCR1 were amplified using specific primers (see Table I⇑) and subjected to agarose gel electrophoresis under certain culture conditions. Band densities were evaluated by volume integration. The data were normalized on the GAPDH mRNA expression.
To clarify the role of IL-8 in suppressing CXCR1 expression in FCS-containing medium, we investigated the effect of IL-8-neutralizing Ab on the [Ca2+]i response to IL-8. After adding IL-8-neutralizing Ab (10 μg/ml; n = 8) for 24 h to podocytes grown in FCS-containing medium, IL-8 did induce an increase of [Ca2+]i. Fig. 5⇑E summarizes the data of the effect of neutralizing IL-8 Ab on IL-8-induced increase of [Ca2+]i.
To investigate whether an up-regulation of CXCR1 takes place at the transcriptional level, we performed a series of semiquantitative RT-PCRs under different cell culture conditions: FCS starvation caused a slight increase of mRNA expression of CXCR1 but the addition of IL-8 in FCS-free medium or the addition of IL-8-neutralizing Ab in FCS-containing medium did not result in a change of mRNA expression of CXCR1 (Fig. 5⇑F). These results indicate that functional CXCR1 down-regulation in the presence of FCS is not regulated on the transcriptional level but due to ligand-induced receptor internalization.
FCS starvation might have also influenced the [Ca2+]i response to the other CXCR agonists. To clarify this point, we assessed the [Ca2+]i response to MIG (0.5 μM, n = 5), SDF-1α (10 nM, n = 5), and BCA-1 (100 nM, n = 5) in FCS-starved cells. When compared with the podocytes grown in FCS, the [Ca2+]i response to the agonists was not significantly different from that of FCS-starved podocytes (data not shown).
CCR and CXCR ligands increase O2− production in podocytes
Activation of the NADPH-oxidoreductase enzyme complex leading to generation of reactive oxygen species has been demonstrated in podocytes in vivo and in vitro. It has been suggested that this generation of reactive oxygen species plays a major role in the pathogenesis of proteinuria (22). The addition of NADPH (0.1 mM) to control podocytes led to a small increase of superoxide anion production (data not shown). After the pretreatment of podocytes with IL-8 (50 nM, n = 5), MIG (0.5 μM, n = 8), SDF-1α (10 nM, n = 6), and BCA-1 (100 nM, n = 4), the addition of NADPH caused a significant increase of superoxide production in comparison to control conditions by 262 ± 42%, 225 ± 56%, 228 ± 40%, and 216 ± 61%, respectively (Fig. 6⇓C). Before stimulation with IL-8, cells were serum starved for 24 h to up-regulate CXCR1.
Effects of CCR and CXCR ligands on NADPH-mediated superoxide anion production in human podocytes. A, A representative original recording of the time course and magnitude of NADPH-oxidase activation after stimulation with chemokine ligands, in this case SDF-1α (10 nM). B, Control cells and cells that had been treated with the respective CXC chemokine for 4 h. C, Control cells and cells that had been treated with the respective CCR chemokine for 4 h. Superoxide generation by NADPH-oxidase activity was calculated by integrating the total counts during the first 15 min and superoxide generation was expressed as nanomoles of O2− generated per milligram of cellular protein per minute.
Like the CXCR ligands, the CCR ligands MDC (100 nM), I-309 (10 nM), TECK (100 nM), and CCL28 (2 μg/ml) caused a significant but smaller increase of superoxide production in comparison to control conditions by 136 ± 9%, 162 ± 7%, 129 ± 8%, and 129 ± 9%, respectively (Fig. 6⇑B).
Superoxide generation by NADPH-oxidase activity was calculated by integrating the total counts during the first 15 min. Superoxide generation was expressed as nanomoles of O2− generated per milligram of cellular protein per minute. Fig. 6⇑A shows a representative original recording of the time course and magnitude of NADPH-oxidase activation after stimulation with CXCR ligands, in this case, SDF-1α (10 nM).
Immunofluorescence studies show the expression of CXCR1, CXCR3, and CXCR5 in kidneys with membranous nephropathy but not in control kidneys
We examined cryostat sections of normal kidney cortex and cortex from two patients with MGN and nephrectomy. The diagnosis of membranous nephropathy was made on kidney biopsies with the typical findings of light, electron, and immunofluorescence microscopy (data not shown).
In control kidneys, there was no or only slight staining of CXCR1–CXCR5 (Fig. 7⇓, A, C, E, G, and I), whereas in MGN a positive fluorescence staining for CXCR1 (Fig. 7⇓B), CXCR3 (Fig. 7⇓F), and CXCR5 (Fig. 7⇓J) could be detected in podocytes (arrows indicate podocytes, Fig. 7⇓). Glomerular CXCR2 (Fig. 7⇓, C and D) and CXCR4 (Fig. 7⇓, G and H) were not detectable in these biopsies (Fig. 7⇓).
Immunofluorescence studies on the expression of CXCR1–5 in control kidney (A, C, E, G, and I) and in a kidney with membranous nephropathy (B, D, F, H, and J). Note that there is no or slight staining of CXCR1 (A), CXCR3 (E), and CXCR5 (I) in the control kidney, whereas in membranous nephropathy a positive fluorescence staining of these CXCRs (B, F, and J) could be detected in podocytes (arrows indicate podocytes).
Discussion
The CC and CXC chemokines are important chemotactic molecules that control leukocyte trafficking and function. Recent studies have shown that these molecules also play an important role in several additional biological functions, such as regulation of lymphocyte development, expression of adhesion molecules, cell proliferation, angiogenesis, virus-target cell interactions, and in various aspects of cancer (23). The CCRs and CXCRs are expressed in a cell-type specific manner in subsets of leukocytes but also in some nonhemopoietic cells, such as endothelial and epithelial cells; e.g., CCR8 have recently been detected in human endothelial cells in culture and in endothelial cells in atherosclerotic plaques (24). In the present study, we investigated the expression of CCRs and CXCRs in human podocytes. Using RT-PCR, we demonstrate glomerular mRNA expression of CCR1–10 and CXCR1–5. The presence of all of the investigated CCRs and CXCRs in the glomerulum is best explained by the presence of several different glomerular cell types including podocytes, endothelial cells, mesangial cells, and macrophages that have invaded the mesangium (25). The RT-PCR technique did therefore not permit an assignment of chemokine receptor expression to specific cell types. In different cell lines of human cultured podocytes, RT-PCR studies showed mRNA expression of the CCR subtypes CCR4, CCR8, CCR9, and CCR10. The ligands for these receptors increase [Ca2+]i in podocytes, indicating that the CCRs are functionally active.
In podocytes grown in FCS, mRNA of the CXCR subtypes 1, 3, 4, and 5 were detected. Under these conditions, CXCR3–5 were functionally active as was demonstrated by the chemokine-induced increases in [Ca2+]i. The half maximal concentrations fit well with those described in other cell types such as lymphocytes (26, 27, 28). It was only by using highly sensitive and stringent RT-PCR conditions that mRNA for CXCR1 could be detected in podocytes. Immunhistochemically, CXCR1 could not be detected in podocytes grown in a medium with FCS. Also, IL-8, a ligand of CXCR1 and CXCR2, did not induce a [Ca2+]i response in podocytes, suggesting that podocytes do not express functional CXCR1 under these conditions. However, we showed that FCS stimulated an IL-8 release in podocytes. In the absence of FCS, IL-8 release was inhibited and up-regulation of functional CXCR1 but not CXCR2 was induced; i.e., CXCR1 could be detected in immunnolabeling studies and IL-8 induced a [Ca2+]i response. The data suggest that FCS-mediated release of IL-8 suppressed the expression of functional CXCR1 in podocytes. This hypothesis was supported in experiments, in which IL-8 increased [Ca2+]i after a 24-h incubation of podocytes with FCS and IL-8 Abs to neutralize the released IL-8.
The different culture conditions used did not significantly influence mRNA expression of CXCR1, indicating that CXCR1 is down-regulated by ligand-induced receptor internalization. IL-8 is known to be a neutrophil chemotactic factor which promotes leukocyte adhesion and leads to the recruitment of polymorphonuclear leukocytes to sites of tissue inflammation (29). IL-8 protein expression along the capillary walls and in the mesangium has been reported in several forms of glomerulonephritis. It has been suggested that IL-8 may be partially responsible for the infiltration of leukocytes during glomerulonephritis (30, 31). In addition, IL-8 might induce albuminuria by altering the metabolism of the sulfated compounds of the glomerular basement membrane (32). In a model of immune complex nephritis, IL-8 was expressed in affected glomerula and injection of an anti-IL-8 Ab reduced glomerular neutrophil infiltration, prevented fusion of foot processes of podocytes, and inhibited proteinuria (33). IL-8 release of podocytes during proteinuric diseases may therefore, in an autocrine fashion, participate in the development of proteinuria in glomerular diseases. mRNA could be detected in podocytes for IL-8 but not for IP-10, SDF-1α, and BCA-1. However, these and other ligands can be released from leukocytes and lymphocytes, infiltrating glomerula in the case of inflammation. In addition, some chemokines are also produced by resident glomerular cells. Synthesis of MCP-1, IL-8, and IP-10 has been demonstrated in mesangial cells (34).
How might activation of CCRs and CXCRs influence podocyte function? In Heymann nephritis, proteinuria is dependent on Ab-induced formation of the complement C5b-9 membrane attack complex. It has been demonstrated that sublytic C5b-9 attack on podocytes causes up-regulation of expression of the NADPH-oxidoreductase enzyme complex by podocytes, which is then translocated to their cell surfaces. Subsequently, reactive oxygen species are produced locally which reach the glomerular basement membrane matrix. Reactive oxygen species initiate lipid peroxidation and subsequent degradation of glomerular basement membrane collagen IV, leading to proteinuria (35, 36). In the present study, we show that CCR and CXCR ligands induce the activation of NADPH-oxidase, a major source for the production of superoxide anions in podocytes. Activation of chemokine receptors in podocytes may therefore contribute to the pathogenesis of proteinuria via activation of NADPH-oxidases, leading to a release of superoxide anions.
In contrast to the ample information about the release of chemokines during renal injury, little is known about the expression of chemokine receptors on glomerular cells in vivo. To our knowledge, only CXCR3 have been detected in mesangial cells of patients with IgA nephropathy, membranoproliferative glomerulonephritis, and rapidly progressive glomerulonephritis, indicating that CXCR3 might contribute to mesangial cell proliferation in these diseases (37). It is well known that cultured cells frequently reflect pathological conditions rather than a physiological situation. With this in mind, chemokine receptor expression may be the result of cell culture conditions. For example, CXCR3 were shown to be functionally active in cultured human mesangial cells. However, slight or no signal of immunological staining for CXCR3 could be detected in normal glomerula, whereas a positive signal was seen in injured glomerula, suggesting that CXCR3 expression is up-regulated during glomerular inflammation (37). In agreement with these findings, only a slight immunohistological staining for CXCRs could be detected in normal human glomerula. This slight and diffuse signal was seen for CXCR3 and CXCR5 in healthy human glomerula. However in patients with MGN, a positive staining for CXCR1, CXCR3, and CXCR5 was detected in podocytes, but not for CXCR2 or CXCR4. With regard to the release of superoxide anions after activation of these receptors in podocytes in culture, these receptors may be involved in the pathogenesis of podocyte injury and proteinuria in MGN. The difference in expression of CXCR4 in cultured podocytes and in podocytes during MGN (podocytes in culture possess CXCR4, whereas during MGN podocytes do not express CXCR4) might be due to the following: 1) CXCR4 might not be present in MGN but instead in other glomerular diseases associated with injury of the podocyte. 2) CXCR4 might only be found under cell culture conditions.
The proinflammatory cytokine IL-1β, which has been shown to induce chemokine receptor expression in astroglioma cells (18), did not stimulate mRNA expression for those CCRs and CXCRs which have not been detected in unstimulated podocytes. This suggests that no additional chemokine receptor expression can be induced by Il-1β in podocytes.
In comparison to the strong staining of the CXCR1 and CXCR5 in podocytes in MGN, CXCR3 staining was weaker. In a recently performed semiquantitative evaluation of immunohistochemical staining for the CXCR3, Romagnani et al. (37) found only a slight positive staining for CXCR3 in two of four kidneys from patients with MGN, a finding that was not different from normal kidneys. The difference in our present study, where a small increase of CXCR3 staining could be observed in podocytes from patients with MGN, might be explained by methodical differences; i.e., in the study by Romagnani et al. (37), paraformaldehyde was used to fix kidney tissue and the avidin biotin-peroxidase method was used to detect Ab reaction. In addition, it might be difficult to quantify positive staining in podocytes. Podocytes represent only 5% of all glomerular cells and are sometimes difficult to localize.
Is there any further evidence that CXC chemokines may participate in the development of glomerular damage and proteinuria? It was recently demonstrated that IP-10 expression is up-regulated in resident glomerular cells in vitro and in vivo in Adriamycin nephropathy. Levels of glomerular IP-10 mRNA expression and glomerular and tubulointerstitial IP-10 protein coincidenced with maximal proteinuria (38). Therefore, IP-10 may participate in the pathogenesis of podocyte injury and proteinuria during Adriamycin nephrosis. In addition to their possible role in glomerular inflammation, the chemokine receptors CCR5 and CXCR4 are also known to serve as coreceptors for the cellular entry of HIV-1 (39). Moreover, CXCR5 has been demonstrated to act as a specific coreceptor for HIV-2 (40). The podocyte plays a crucial role in the pathogenesis of HIV nephropathy and it has recently been shown that the loss of podocyte maturity markers and podocyte proliferation occur in HIV-associated nephropathy (41). Very recently, it was demonstrated that CCR5 and CXCR4 were not expressed in resident glomerular cells in HIV-associated renal disease (42). The possible role of CXCR5 in HIV-2-induced nephropathy has to be investigated in future studies. Expression of CXCR5 was primarily demonstrated on B lymphocytes (43). Interestingly, we show here that CXCR5 is also expressed on a nonhemopoietic cell type; the podocyte. Several earlier studies have shown that podocytes are able to express other leukocyte-associated markers, such as CD2-associated protein or CD68 under normal and pathological conditions, respectively, indicating that there are some similarities between protein expression of leukocytes and podocytes (44, 45). Recently, it has also been shown that injured podocytes undergo a process of transdifferentiation with the acquisition of epitopes that are characteristic of activated macrophages (46). The presence of distinct chemokine receptors on podocytes further supports the idea that podocytes can turn into inflammatory cells which are actively involved in the inflammatory processes that occur in glomerular diseases.
In conclusion, cultured podocytes express functional CCR and CXCR receptors. Activation of these receptors may contribute to podocyte damage and proteinuria during glomerular diseases. Therefore, it will be interesting to find out whether chemokine receptor antagonists can reduce proteinuria in patients with inflammatory glomerulopathy.
Acknowledgments
We thank Temel Kilic, Charlotte Hupfer, Monika von Hofer, Barbara Müller, and Petra Daemisch for their excellent technical assistance.
Footnotes
-
↵1 This work was supported by the Else-Kröner Fresenius Stiftung and Fonds zur fòrderung der Wissenschuftlichen Forschung, SFB 07, Project 07 (to D.K).
-
↵2 T.B.H. and H.C.R. contributed equally to this manuscript.
-
↵3 Address correspondence and reprint requests to Dr. Hermann Pavenstädt, Medizinische Universitätsklinik Freiburg, Abteilung IV, Labor C4, Hugstetter Strasse 55, D-79106 Freiburg, Germany. E-mail address: paven{at}mm41.ukl.uni-freiburg.de
-
↵4 Abbreviations used in this paper: MGN, membranous nephropathy; IP-10, IFN-inducible protein; MCP, monocyte chemoattractant protein; WT1, Wilm’s tumor Ag; [Ca2+]i, intracellular calcium; MIG, monokine induced by IFN-γ; SDF-1α, stromal cell-derived factor 1α; BCA-1, B cell-attracting chemokine 1; MDC, macrophage-derived chemokine; TECK, thymus-expressed chemokine; CCL28, C chemokine ligand 28; HCC, hemofiltrate CC chemokine.
- Received February 2, 2002.
- Accepted April 19, 2002.
- Copyright © 2002 by The American Association of Immunologists