Roles of SLC/CCL21 and CCR7 in Human Kidney for Mesangial Proliferation, Migration, Apoptosis, and Tissue Homeostasis¹

Chemokines were initially identified by their ability to induce the migration of leukocytes, but their biological roles have subsequently expanded (1). The chemokine superfamily is separated into four general branches (CL, CCL, CXCL, and CX3CL chemokines) based upon cysteine residues within their primary amino acid sequence. The biological actions of chemokines are mediated through their interaction with a large family of seven-transmembrane-spanning G protein-coupled receptors (i.e., C, CC, CXC, and CX3C receptors) (2).

Chemokines and chemokine receptors have been further classified according to their function and regulation of expression. Some chemokines are important in the control of inflammatory processes (3). During inflammatory glomerular disease or transplant rejection, chemokines such as monocyte chemoattractant protein (MCP)³-1/CCL2 and RANTES/CCL5 are up-regulated by both intrinsic renal cells and infiltrating leukocytes (4).

An additional subgroup of chemokines is involved in homeostasis of lymphocyte and dendritic cell trafficking during immune surveillance. The regulated expression of the chemokine secondary lymphoid tissue chemokine (SLC/CCL21) and its corresponding receptor, CCR7, represents a prototypic model for chemokine/chemokine receptor functions in normal lymphoid tissue (5, 6). SLC/CCL21 is constitutively produced by high endothelial venules and by stromal cells within T cell zones of lymphoid organs. CCR7 is present on T cell subpopulations (7). CCR7 is up-regulated by maturing dendritic cells and helps direct the dendritic cells to secondary lymphoid organs (8). A spontaneous mouse line lacking SLC/CCL21 (9) and CCR7 knockout mice (10) shows impaired homing of T cells into lymph nodes and Peyer’s patches within the small intestine.

The biology of chemokines and chemokine receptors has become more complex with the demonstration of functional chemokine receptor expression on nonhematopoietic tissues. For example, endothelial cells (11), epithelial cells (12), microglial cells (13), neurons (14), and mesangial cells (MC) (15, 16) have been shown to express chemokine receptors. The role of chemokine receptor expression by nonhematopoietic cells is not completely understood, but it is thought to be important in angiogenesis and tissue remodeling, e.g., during atherosclerosis, wound healing, and tumor metastasis (17, 18).

MC are involved in local immune regulation within the glomerulus (19). We have previously described the inducible expression of the chemokine receptor CCR1 by human MC (15). In this work we describe the functional and constitutive expression of CCR7 on human MC. Interestingly, the CCR7 ligand SLC/CCL21 is constitutively expressed by glomerular podocytes in human kidney. Effects of SLC/CCL21 on mesangial migration, proliferation, and cell death were found, suggesting a potential homeostatic function for this chemokine/chemokine receptor pair in the glomerulus.

Human tissue was used following the guidelines of the Ethics Committee of the Medical Faculty of the University of Heidelberg (Heidelberg, Germany). Adult kidneys removed for reflux nephropathy were used to establish specific staining protocols for each Ab. Immunohistochemistry was conducted on frozen tissue sections of normal human kidney essentially as described previously (20) using an avidin biotinylated enzyme complex method. Staining for CCR7 was performed using a specific mouse mAb, as described (21). mAbs specific for human SLC/CCL21 and CD31/platelet endothelial cell adhesion molecule (PECAM)-1 were obtained from R&D Systems (Wiesbaden, Germany). Control experiments entailed immunohistology with nonimmune mouse or rat IgG, respectively, and without primary Ab. Negative controls for the avidin biotinylated enzyme complex stain generated with nonimmune control Ab and without primary Ab did not show any staining (data not shown).

Immortalized human MC were grown as described previously (15). This MC line showed no dedifferentiation within ∼100 passages during a 36-mo cultivation period. Furthermore, it was characterized for antigenic markers typically expressed by MC in vivo and in vitro and is known to have a specific expression pattern for chemokines and the chemokine receptor CCR1 identical to primary human MC (15). For all experiments, passages 51–65 were used. Different preparations of primary human MC served as controls and were cultured as previously published (19).

Human MC were stimulated for 4–96 h with human rTNF-α (20 ng/ml), human rIL-1β (2 ng/ml), and human rIFN-γ (10 ng/ml), either alone or in combination. Additional stimuli tested included TGF-β-1, -2, and -3, IL-4, IL-12, RANTES/CCL5, MCP-1/CCL2, eotaxin/CCL11, monokine induced by IFN-γ/CXCL9, IFN-γ-inducible protein-10/CXCL10, and SLC/CCL21 (all from R&D Systems). Control cells were kept under standard conditions. Cells were harvested by trypsinization and total RNA was prepared as described (22).

Semiquantitative PCR analysis was done as described (15). For real-time quantitative RT-PCR, 2 μg of isolated total RNA underwent random primed reverse transcription using a modified Moloney murine leukemia virus reverse transcriptase (Superscript; Life Technologies, Karlsruhe, Germany). In parallel, 2-μg aliquots were processed without reverse transcription to control for contaminating genomic DNA. Glomeruli were obtained from unaffected kidney compartments from tumor nephrectomies by manual microdissection. Microdissection was verified by demonstration of a podocyte-specific gene (Wilms-tumor gene-1) expressed specifically within the glomerular compartment. Real time RT-PCR was performed on a TaqMan ABI 7700 Sequence Detection System (PE Applied Biosystems, Weiterstadt, Germany). Thermal cycler conditions contained holds at 50°C for 2 min and 95°C for 10 min followed by 40 cycles of 95°C for 15 s and 60°C for 60 s. GAPDH was used as reference gene. All controls consisting of ddH₂O were negative for target and housekeeper.

The following oligonucleotide primers (300 nM) and probes (100 nM) were used: human CCR7 (GenBank accession no. NM 001838; bp 816–878) sense (5′-GAGGCCCAGAGGATCGCT-3′) and antisense (5′-ACTTGGAGTTGATGATTTGCGG-3′), internal fluorescence-labeled probe (FAM) (5′-CAACCACATCAAGCTGTCGGGCAG-3′), human SLC/CCL21 (GenBank accession no. AB 002409; bp 194–275) sense (5′-CGCAGCTACCGGAAGCAG-3′) and antisense (5′-CTGCCTGAGAGCGCTTGC-3′), and probe (FAM) (5′-CTCCATCCCAGCTATCCTGTTCTTGCC-3′). SLC primer pairs are intron-spanning leading to exclusive amplification of cDNA. For CCR7 no cDNA-specific primers were available. Therefore, non-reverse-transcribed samples were analyzed in parallel to control for contaminating genomic DNA. All primers and probes were obtained from PE Applied Biosystems.

Template sets for human chemokine receptors for use in RNase protection assays were obtained from BD PharMingen (San Diego, CA). RNase protection assays were performed according to the manufacturer’s instructions using 20 μg of total RNA from human MC to analyze the expression of CCR7 mRNA. Equal amounts of tRNA were used as control to exclude incomplete digestion of the probes. After drying, gels were exposed on phosphor screens for use with a Storm 840 PhosphorImager (Molecular Dynamics, Sunnyvale, CA).

For FACS analysis human MC were detached with PBS/10 mM EDTA (pH 8) and stained for CCR7 using a specific mouse mAb and a two-step amplification method as described previously (21). The CCR7 signal was analyzed using a FACScan flow cytometer (BD Biosciences, Mountain View, CA). Appropriate IgG isotype preparations were used to control for unspecific staining.

To assess the proliferative activity of human MC, MTT (Sigma-Aldrich, Deisenhofen, Germany) assays were performed (23). Aliquots of 20 × 10³ cells in 100 μl medium were cultured in 96-well microtiter plates for 24 h under standard conditions to yield firmly attached and stably growing cells. After discarding the supernatants, 50 μl of medium containing chemokines and/or cytokines were added and the cells were incubated from 24 to 72 h. Then 50 μl of a 1 mg/ml solution of MTT were added. After a 3-h incubation at 37°C formazan crystals were dissolved by addition 50 μl isopropanol. Absorbance was measured at 550 nm using a DYNATECH MR7000 ELISA reader. For each experiment at least six wells were analyzed per experimental condition and time point.

In chemotaxis assays 2 × 10⁵ MC per well were plated onto Transwell filter inserts (8-μm pore size; Costar, Cambridge, MA) coated with fibronectin (Boehringer Mannheim, Mannheim, Germany). Assay medium consisted of DMEM, 0.1% BSA (Sigma-Aldrich), and 10 mM HEPES (Life Technologies). After a 3-h incubation at 37°C, chemotactic factors (diluted in assay medium) were added to 12-well tissue culture plates. Transwells were inserted to the wells and then incubated for 4 h. Cells that had migrated to the bottom of the filter and/or the bottom chamber were collected by trypsinization and counted using flow cytometry as described previously (15).

Apoptosis of MC was induced by Fas/CD95 ligation according to the method of Gonzalez et al. (24) after starvation of the cells in serum-free medium and prestimulation with IFN-γ to induce Fas/CD95 surface expression. For analysis of chemokine effects, cells were pretreated with 250 ng/ml SLC/CCL21 before adding the activating anti-human Fas Ab (Biomol, Hamburg, Germany). Three different assays were used to study apoptosis. Flow cytometric cell cycle analysis using propidium iodide staining was performed as described (25). For visualization of chromatin fragmentation, MC were seeded on chamber slides (Nunc, Wiesbaden, Germany). After treatment with test substances cells were fixed with ethanol and stained with the nuclear dye Hoechst 33258 (Hoechst, Bad Soden, Germany) 5 μmol/ml. The percentage of apoptotic cells was determined by immunofluorescence microscopy, counting nuclei with condensed or fragmented chromatin. Three independent experiments were performed; at least 300 cells were analyzed per condition. Counting was performed in a blinded manner by two investigators. For measurement of caspase-3 activity (26), a commercial assay (R&D Systems) was used according to the manufacturer’s specifications. After induction of apoptosis as described above, the human MC were lysed and caspase-3-specific proteolytic activity was quantitated spectrophotometrically. Three experiments were done analyzing duplicates for any experimental condition.

Scratch assays (27, 28) were performed as an in vitro model for “wound healing.” Human MC were grown under standard conditions in six-well tissue culture dishes until confluent. At the time point 0 h the cell monolayer was scratched in a standardized manner with a plastic spatulum to create cell-free areas (2 mm in width) in each well. Afterward the cell culture medium was replaced by test medium containing either standard medium or medium supplemented with human rSLC/CCL21 (250 ng/ml). The “wounds” were observed at 6-h intervals for 48 h by phase contrast microscopy and documented by photography. Each experiment was performed in duplicate analyzing at least four scratches per well.

Values are provided as mean ± SEM. Statistical analysis was performed by unpaired t test. Significant differences (see Figs. 4–6) are indicated for p values <0.05 (∗) or 0.01 (∗∗), respectively.

Representative immunohistochemical sections of human kidney stained with an anti-SLC/CCL21 Ab are shown in Fig. 1,A. Normal adult kidney showed a strong staining in the periphery of capillaries in a podocyte-like pattern. Staining for the corresponding receptor CCR7 was seen primarily in the mesangium but also in medial artery smooth muscle cells (Fig. 1,B). Expression of SLC/CCL21 and/or CCR7 by glomerular endothelial cells could be excluded by a specific endothelial counterstain (CD31/PECAM-1). Data shown in Fig. 1 are representative for kidney tissue prepared from seven adult individuals of both female and male gender.

The immunohistochemistry suggested a constitutive expression of SLC/CCL21 by normal human podocytes. Because differentiated human podocyte cell lines are not available, the expression of SLC/CCL21 was evaluated in human glomeruli isolated from normal adult kidney using manual microdissection. Real-time PCR analysis of the microdissected glomeruli revealed expression of SLC/CCL21 mRNA (Fig. 2). A positive signal for the Wilms-tumor gene-1 served as control for the presence of podocyte-specific transcripts within this RNA preparation. Sequencing of the PCR product proved the specificity of glomerular SLC/CCL21 mRNA expression (data not shown). Cultured human MC and human microvascular endothelial cells failed to express SLC/CCL21 mRNA, suggesting glomerular epithelial cells as the source of SLC/CCL21 mRNA (data not shown).

Because glomerular staining for CCR7 resulted in a mesangial pattern, we examined cultured human MC for expression of CCR7. RNA was prepared from cells growing under standard conditions as well as from cells that had been stimulated with a combination of the cytokines TNF-α, IL-1β, and IFN-γ to simulate a proinflammatory situation. By RT-PCR, specific products for CCR7 mRNA were amplified from both unstimulated and stimulated cells. Equivalent results were obtained with immortalized MC and three different preparations of primary human MC (Fig. 3, A and B). To confirm and quantitate the CCR7 expression, RNase protection assays were performed. A constitutive expression of CCR7 was found analogous to the RT-PCR experiments (Fig. 3 C).

Because both semiquantitative RT-PCR and RNase protection assay suggested a constitutive expression of CCR7 by human MC, a more comprehensive expression analysis was performed using real-time PCR. RNA was prepared from human MC stimulated with FBS, various cytokines (IFN-γ, IL-1β, IL-4, IL-5, IL-10, IL-12, TNF-α, TGF-β1, TGF-β2, and TGF-β3), or chemokines (MCP-1/CCL2, RANTES/CCL5, eotaxin/CCL11, ELC/CCL19, and SLC/CCL21) from 4 to 96 h. No significant modulation of CCR7 mRNA expression was found under any of these conditions (data not shown).

To confirm expression of CCR7 protein on human MC, FACS analysis was conducted with mAbs specifically directed against this receptor (10). As demonstrated in Fig. 3,D, the human MC line showed a low but reproducible staining for CCR7. Consistent with the mRNA data, the amount of receptor detected by flow cytometry did not differ when MC cells were stimulated with cytokines (data not shown). Primary human MC also showed a comparable staining by FACS analysis for CCR7 (Fig. 3 E).

To demonstrate that CCR7 mediates MC chemotaxis, migration assays were performed using a Transwell filter system. SLC/CCL21 induced a directed migration of human MC in a dose-dependent fashion (Fig. 4). Comparable results were obtained in three independent sets of experiments, each conducted in triplicate.

CCR7 activation has an influence on the proliferative activity of MC. As shown in Fig. 5,A, stimulation with the CCR7 ligand SLC/CCL21 increased proliferation of human MC in a dose-dependent manner in a range from 10 to 250 ng/ml. To determine optimal stimulation time, human MC were incubated with SLC/CCL21 up to 72 h. The highest pro-proliferative effect of SLC/CCL21 was seen when human MC were stimulated for 48 h (Fig. 5 B).

The potential role of CCR7 for MC survival was studied using Fas/CD95-mediated cell death assays. Three methods were used to study the influence of SLC on MC apoptosis: cell cycle analysis, Hoechst staining, and caspase-3 activity. Cell cycle analysis revealed a background of 7.8 ± 0.6% apoptotic cells under normal conditions. After serum starvation and IFN-γ stimulation human MC expressed surface Fas/CD95 (data not shown), and 24.5 ± 3.7% of the cells belonged to a population with a subG₁ DNA content (Fig. 6,A). Fas ligation increased the amount of MC displaying a subG₁ DNA content to 52.4 ± 3%, consistent with a marked increase in apoptosis (Fig. 6,B). When MC were prestimulated with SLC/CCL21 before induction of cell death the percentage of apoptotic cells was reduced markedly to 30.8 ± 4.6% (Fig. 6,C). Staining with Hoechst dye was used to visualize cells with fragmented chromatin. Under control conditions 10.7 ± 1.9% of the cells were found to be apoptotic, and after serum starvation and IFN-γ stimulation 16.5 ± 4.5% of the cells were found to be apoptotic. Subsequent Fas ligation induced apoptosis in 45.2 ± 5.8% of human MC. Prestimulation with SLC/CCL21 reduced Fas-induced cell death effectively to 20.3 ± 7.6% (Fig. 6,D). In addition, SLC/CCL21 affected activation of caspases. Caspase-3 activity was found to be increased 2.8-fold upon Fas ligation of MC compared with control conditions. Coincubation with SLC/CCL21 reduced caspase-3 activity significantly (Fig. 6 E).

Finally, signaling through CCR7 was tested in an in vitro “wound healing” model system that requires both migration and proliferation of cells. Defined lesions were generated in subconfluent layers of human MC and the repopulation of denuded areas was studied. Under standard conditions no relevant change of the gap was apparent 12 h after inducing the lesion. After 24 h the gap started to close but was still visible at 48 h (Fig. 7, left panels). In contrast, upon addition of SLC/CCL21 accelerated wound closure was already noticeable after 12 h. At 24 h the lesions were almost closed and at 48 h the previously cell-free areas were completely repopulated (Fig. 7, right panels).

A balance between endothelial, mesangial, and visceral epithelial cells (e.g., podocytes) and their extracellular matrix represents the prerequisite for the structural and functional integrity of the glomerulus. During glomerular injury, both podocyte function and morphology are altered. This frequently coincides with MC proliferation, cell death, glomerular scarring, and sclerosis. MC proliferation has been studied in diverse model systems of glomerular disease and is found to be regulated by a large number of growth factors and mediators, including platelet derived growth factor (28), basic fibroblast growth factor (29), and insulin-like growth factors (30).

In addition to proliferation, the directed migration of cells is necessary for tissue remodeling after injury. In vitro migratory activity of MC has been demonstrated in response to growth factors (31) and angiotensin II (32). During embryonal glomerulogenesis MC migration from the vascular pole into the glomerulus is thought to represent an important step (33, 34). In a rat nephritis model Hugo et al. (35) demonstrated that proliferation is only one mechanism by which MC repopulate the glomerular space after induction of glomerular damage and mesangiolysis. The second important phenomenon is a directed migration of “reserve” MC from the juxtaglomerular apparatus into the glomeruli. In this context it is possible that the local synthesis of chemokines and activation of their receptors may help control MC migration.

It has been suggested that chemokines and chemokine receptors expressed by intrinsic renal cells play a role in kidney development and in the cellular homeostasis of the adult organ (36). In the present study we analyzed the expression of CCR7 and its ligand secondary lymphoid tissue chemokine (SLC/CCL21) in human renal tissue. Immunohistochemistry of human kidney showed a clear staining pattern for CCR7 on MC and SLC/CCL21 on podocytes. These findings were confirmed in cultured cells and isolated glomeruli. Functional assays revealed an influence of SLC/CCL21 on mesangial migration, proliferation, cell death, and in vitro “wound healing.” Interestingly, the positive influence of SLC/CCL21 on mesangial proliferation is different from the effect on blood cells. The proliferation rate of normal hematopoietic or leukemia progenitor cells was reduced upon stimulation with SLC/CCL21 (37), suggesting differential actions of this chemokine on hematopoietic or nonhematopoietic cells.

In addition to proliferation, tissue homeostasis including that of MC is regulated by apoptosis (38, 39, 40, 41). In this work we describe that SLC/CCL21 acts as an antiapoptotic factor for MC. As the MC is a specialized pericyte, chemokines and their receptors may function to control vascular smooth muscle homeostasis via proliferation and prevention of apoptosis. This novel role of the pair SLC/CCL21 and CCR7 is reminiscent of the regulation of apoptosis via CX3CR1 described by Boehme et al. (13) for brain microglia. Analogous to the protective effect of SLC/CCL21 on MC survival described in this work, the CX3CR1 ligand fractalkine was shown to maintain cell survival and inhibit Fas ligand-induced apoptosis in brain cells. Mediating cell death and survival could be an important function of chemokine/chemokine receptor interaction, especially on nonimmune cells.

The constitutive expression of SLC/CCL21 by podocytes and of CCR7 on MC is of interest, because other chemokines and chemokine receptors are generally expressed after cytokine stimulation. For example, we described the induction of functionally active CCR1 on human MC after stimulation with a combination of TNF-α, IL-1β, and IFN-γ (15). Furthermore, CCR1 is up-regulated in kidney cortex of mice with immune complex glomerulonephritis (42). Romagnani et al. (16) reported mesangial expression of CXCR3 in biopsies from patients with mesangioproliferative glomerulonephritis. Therefore, a role for CCR1 and CXCR3 in inflammatory glomerular disease has been proposed.

In this context a potential mesangial expression of CXCR3 is of special interest, because in mice this receptor may bind not only IFN-γ-inducible protein-10/CXCL10 and monokine induced by IFN-γ/CXCL9 but also SLC/CCL21 (43). Whether this also applies to the human system remains in question (44, 45). In any case, in mice SLC/CCL21 generated by podocytes may bind to both CCR7 and CXCR3. This redundancy of SLC/CCL21 receptors in mice may also explain why CCR7 knockout mice do not show a glomerular phenotype.

The finding of a paired expression of a potential growth factor and its receptor on different but adjacent intrinsic glomerular cells as shown in this work is not unique. For example, vascular endothelial growth factor (VEGF), a potent mitogen for endothelial growth (46), is expressed by podocytes (20, 47) while the respective VEGF receptors are expressed on glomerular endothelial cells and MC (48, 49). Proliferation of human MC occurs upon stimulation with recombinant VEGF (49). To this end, SLC/CCL21 could represent a factor that helps regulate the cellular homeostasis between podocytes and MC, including their number, migration, and interaction with the intervening basement membrane.

The complexity and apparent redundancy of chemokine biology provides a high degree of effectiveness and flexibility in vivo. This redundancy may also exist with regard to tissue homeostasis. Although CCR7 knockout mice have no obvious renal phenotype under basal conditions, the significance of the deleted gene during experimental stress will have to be evaluated. Indeed, it has been shown that genes important in renal development and physiology do not always show renal phenotypes in knockout animals (50).

Based on these observations it is tempting to speculate that migration, proliferation, and apoptosis of MC during glomerulogenesis and glomerular injury might be controlled by the local synthesis of chemokines by podocytes with activation of their respective receptors on MC. At present such a role for SLC/CCL21 and CCR7 for a homeostatic balance between podocytes and MC remains an interesting hypothesis deserving further investigation.

We thank Monika Fink for her expert technical help.