Journal Logo

Experimental Transplantation

Renal Transplantation: Examination of the Regulation of Chemokine Binding During Acute Rejection

Ali, Simi; Malik, Ghada; Burns, Alice; Robertson, Helen; Kirby, John A.

Author Information
doi: 10.1097/01.TP.0000155961.57664.DB
  • Free

Abstract

The prototypic inflammation associated with cellular rejection of a renal allograft has been associated with specific patterns of immune cell infiltration and accumulation. Indeed, the recruitment of mononuclear cells to the tubular epithelium is regarded as a defining feature of the rejection process (1). Although leukocyte migration can be induced through the actions of complement components, leukotrienes, platelet-activating factor, or other chemotactic agents, the importance of chemokines is currently receiving great attention (2). In particular, the CC subfamily of chemokines, including CCL5 (regulated on activation normal T-cell expressed and secreted [RANTES]), is thought to play an important role in recruiting mononuclear immune cells to the tubular epithelium during acute renal allograft rejection (3).

The sequestration of active chemokines by glycosaminoglycan (GAG) molecules within the extracellular matrix or on the surface of cells is thought to be essential for producing the static concentration gradients of chemokines that are required for directional leukocyte migration (4). This binding process is stabilized by interaction between basic amino acid sequence motifs on chemokine molecules and heavily sulphated, anionic domains within the GAG molecules (4–6). Of the various GAG species, heparan sulphate (HS) is known to play a dominant role in chemokine sequestration (7).

Chemokine binding is important for localized presentation of chemokines on the apical surface of vascular endothelium (8–10) where soluble cytokine molecules would otherwise be displaced by the rapid flow of blood. Indeed, interaction between chemokines and GAGs seems critical for extravasation, because CCL5 molecules with mutant GAG-binding motifs (11) have a restricted ability to induce transendothelial leukocyte migration both in vitro (12) and in vivo (13). Immunocytochemical studies of rejecting renal allografts have shown that CC chemokines, including CCL5, CCL3 (macrophage inflammatory protein-1α), CCL4 (macrophage inflammatory protein-1β), and CCL2 (monocyte chemotactic protein-1), tend to be concentrated on the basolateral aspects of the tubular epithelial cells with less staining of the interstitial tissues (3). Semiquantitative immunofluorescence assays have also provided some evidence for the existence of chemokine concentration gradients within the graft (14); these might provide a vectorial cue for specific recruitment of mononuclear cells to the tubules.

HS is highly variable in terms of its overall chain length, structure, and degree of sulphation (15). During synthesis, a copolymer of N-acetyl-glucosamine and glucuronic acid is attached by a tetrasaccharide linkage to the proteoglycan core protein. Glucosamine transferases commit the extending chain to synthesis of HS rather than chondroitin sulphate. As the chain extends, it is potentially modified by the addition of sulphate residues at the N-acetyl group of glucosamine and at a range of hydroxyl groups (O-sulphation) within the disaccharide unit. In addition, glucuronic acid may be modified by epimerization of carbon 5 to form iduronic acid. The O-sulphation and epimerization modifications are dependent on initial replacement of the N-acetyl group of glucosamine with an N-sulphate group (16); indeed, full activity of the carbon 5 epimerase requires N-sulphation of an adjacent glucosamine (17).

N-sulphation of HS is catalyzed by the N-deacetylase/N-sulphotranferase (NDST) family of enzymes, which contains four human isoforms (NDST-1, 2, 3, and 4) (18). Although these enzymes all share the same basic function, they do appear to have subtly different activities; NDST-3 and 4 are expressed at low levels with a restricted distribution. It has been shown that the proinflammatory cytokines interferon (IFN)γ and tumor necrosis factor (TNF)α can increase the expression of NDST-1 by immortalized endothelial cells (19), providing an explanation for the observation that the sulphation of endothelial GAGs is increased by these cytokines (20). On this basis it is reasonable to propose that cytokine-mediated variations in the distribution and sulphation of HS might provide a further mechanism to regulate the potential of intragraft chemokines to recruit immune cells during the development of inflammation.

The current study was designed to explore the role played by changes in the biology of GAG molecules within kidney sections during the development of episodes of acute renal allograft rejection. Initial studies were focused on assessing changes in the absolute level of expression of the major classes of GAG and on the correlation of these findings with the abundance and spatial distribution of the important chemokine CCL5. The potential for changes in the expression and sulphation of HS to modify CCL5 binding was also investigated in vitro by sensitive modeling with immortalized and primary lines of human vascular endothelial cells and by studying kidney biopsy sections. These cells were chosen for this study because they are representative of the first donor tissues encountered by recipient leukocytes after organ transplantation and because the endothelium is known to regulate the passage of blood-borne leukocytes into the graft tissues.

MATERIALS AND METHODS

Cell Culture

The human microvascular endothelial cell line human mammary epithelial cell (HMEC)-1 (21) (provided by Dr. F. J. Candal, CDC, Atlanta, GA) was cultured in MCDB131 medium (Sigma-Aldrich, Dorset, UK) supplemented with 1 ng/mL hydrocortisone (Sigma-Aldrich), 10 ng/mL human epidermal growth factor (Peprotech, London, UK), and fetal bovine serum (Gibco, Paisley, UK) with penicillin and streptomycin (Sigma-Aldrich). The cells were maintained in 75 cm2 flasks and were routinely subcultured as required.

The human umbilical vein endothelial cell line (HUVEC; PromoCell) was cultured in endothelial cell basal medium (PromoCell) supplemented with 2% fetal calf serum, 0.1 ng/mL human epidermal growth factor, 1 ng/mL human basic fibroblast growth factor, 1 μg/mL hydrocortisone, 50 μg/mL gentamicin, 50 ng/mL amphotericin B, and endothelial cell growth supplement (all from Sigma-Aldrich). The cells were seeded at 5,000/cm2 and incubated at 37°C in 5% CO2.

Antibodies and Labeled Chemokines

Murine monoclonal antibodies (all immunoglobulin [Ig]M) specific for the GAGs HS (10E4 epitope, predominantly N-sulphated domains) (22), chondroitin-4-sulphate (C4S), and chondroitin-6-sulphate (C6S) were all from Seikagaku Corp. (Falmouth, MA). A murine anti-CD31 antibody (DAKO, Ely, UK) was used for positive control labeling of endothelial cells. Polyclonal goat anti-RANTES was obtained from R&D Systems (Abingdon, UK). Primary antibodies were detected by secondary labeling with fluorescein isothiocyanate (FITC)-conjugated goat antimouse (Becton Dickinson, Oxford, UK) or FITC-conjugated rabbit antimouse (DAKO) and rabbit antigoat (Sigma-Aldrich) Igs. All antibodies were applied at a predetermined optimum concentration in phosphate-buffered saline (PBS) containing 3% (v/v) normal goat serum (Sigma-Aldrich). Biotin-conjugated RANTES and FITC-conjugated avidin were both from R&D Systems.

Tissue Samples and Immunofluorescence

Paraffin-embedded archival renal transplant biopsy specimens representing histologically defined acute rejection (Banff ’97 scores of at least i2 and t2) (23) were used in this study. Donor kidneys that were anatomically unsuitable for transplantation provided normal control tissue. All samples were used in accordance with local ethics committee approval.

For single-color immunofluorescence, 5-μm sections were dewaxed, rehydrated, and incubated in 20% (v/v) normal rabbit serum in tris-buffered saline (NRS/TBS) for 2 hr at 4°C. The sections were then incubated with appropriate primary antibodies at 10 μg/mL in NRS/TBS overnight at 4°C. After the sections were washed, rabbit (F[ab′]2) antimouse Igs conjugated to FITC (DAKO) were added for 2 hr at 37°C. The sections were then washed and mounted (fluorescence mounting medium; DAKO). For dual-color labeling the rehydrated sections were first treated for antigen retrieval by microwaving twice in citrate buffer (pH 6) for 5 min. After the sections were blocked with NRS/TBS, they were treated with polyclonal (IgG) goat anti-CCL5 (R&D Systems) and monoclonal anti-HS overnight at 4°C. After the sections were washed, they were treated for 2 hr at 37°C with rabbit antimouse Ig conjugated with TRITC (Sigma-Aldrich) and rabbit antigoat Ig conjugated with FITC (Sigma-Aldrich). The sections were then washed and mounted.

All sections were analyzed with laser scanning confocal microscopy (LSCM). To gain a semiquantitative measure of GAG and chemokine expression at different parts within the biopsy specimens, optical sectioning was performed by scanning across the XY-axis at 1.5-μm increments in the Z axis, and a projected Z series was constructed by integration of serial images throughout the depth of the section. The fluorescence intensity was quantified by measuring the mean pixel intensity within a standard size box drawn in three random areas of the section or by selecting a line across the Z-series image in regions representative of tubular epithelium or endothelium. The image fluorescence intensity was then translated into mean pixel intensity for graphic analysis.

For dual-labeled sections the Leica TCS SP2 UV LCSM was set up to image the fluorochromes sequentially to avoid possible crosstalk between the two emission spectra (FITC excitation 488 nm, emission 520 nm; TRITC excitation 543 nm, emission 578 nm). The areas of interest were selected and imaged with the same setup. Combined maximum projections of the collected Z-series were examined for colocalization of the fluorophores by observing the mixing of the two assigned colors. The threshold for positive staining was based on the level of background fluorescence recorded on control sections treated with normal goat serum instead of primary antibody followed by staining with the secondary antibody. Our group previously demonstrated the reproducibility of data generated using this method with a coefficient variation of less than 12%, determined by repeated staining and examination of three separate biopsy specimens (24).

Immunochemical Investigation of Human Microvascular Endothelial Cell Line (HMEC-1)

The HMEC-1 cell line was grown on four-chambered Permanox slides (Nunc; NE Lab Supplies, Durham, UK) for 48 hr at 37°C until 50% to 60% confluency was achieved. The cells were fixed within the chambers using 4% (w/v) paraformaldehyde for 15 min and washed in sterile PBS for 5 min at room temperature.

Immunohistochemistry was performed by blocking with PBS containing 3% (v/v) normal goat serum (Sigma-Aldrich) for 3 hr at 4°C. The cells in each chamber were incubated with antibodies specific for HS, C4S, or C6S at 10 μg/mL in 3% normal goat serum in PBS and incubated overnight at 4°C. The specificity of the anti-HS antibody was assessed by incubating the HMEC-1 cells with antibody after treatment with 10 mU of heparitinase (Seikagaku Corp) for 3 hr at 37°C. After the cells were washed twice with PBS, they were stained with goat antimouse-FITC at 10 μg/mL in 3% normal goat serum for 3 hr at 4°C. Negative controls included omission of the primary antibody; staining with CD31 antibody was used as a positive control. The detachable chamber structure was removed after incubation with the secondary antibody. The remaining slides were then washed in PBS and allowed to air dry before mounting in fluorescent mount (DAKO). All incubations were performed in humidified chambers to prevent dehydration and minimize nonspecific antibody binding. The preparations were examined using semiquantitative LSCM, and mean pixel intensity values were recorded.

Fluorimetric Measurement of CCL5 Binding

Cell Culture

HMEC-1 cells were grown in four-chambered slides as described in the previous section, and cells in some wells were stimulated for 24 hr with fresh media containing 100 IU/mL of the proinflammatory cytokines IFNγ and/or TNFα (Peprotech). After incubation, the cells were all fixed in 4% paraformaldehyde and washed in PBS. The cells were then covered with 150 μL of PBS containing 10 μL of biotin-conjugated CCL5 (4 μg/mL). After incubation overnight at 4°C, the cells were washed and 150 μL PBS containing 10 μL FITC-avidin (10 μg/mL) was added. After incubation for 3 hr at 4°C, the cells were washed, the chambers were removed, and each preparation was air dried before mounting. Control preparations included omission of biotinylated CCL5 and pretreatment of the endothelial cells with heparitinase to remove HS. Endogenously produced CCL5 was examined by immunofluorescence labeling with polyclonal goat anti-CCL5 (R&D Systems) followed by FITC-conjugated rabbit antigoat IgG (Sigma-Aldrich). All the preparations were examined using semiquantitative LSCM.

Tissue Section

Cryostat sections from normal human kidney and acute rejection biopsies were fixed in acetone, air-dried, and rehydrated in PBS. Endogenous biotin was blocked using a biotin-blocking kit obtained from Vector Laboratories (United Kingdom). Biotinylated recombinant human RANTES or control (FLUOROKINE, R&D Systems) was applied, and the sections were incubated overnight at 4°C. After the sections were gently washed in TBS, they were incubated for 30 min at RT with an avidin biotin alkaline phosphatase complex. Binding of biotinylated RANTES was visualized with Vector Red substrate to which levamisole was added to block endogenous alkaline phosphatase. The sections were analyzed using a Leica DMR microscope with image-grabbing facilities.

Real-Time Polymerase Chain Reaction

Total RNA was isolated from HUVEC cells that had been stimulated with 100 IU/mL of the proinflammatory cytokines IFNγ and TNFα for 24 hr using TRIzol (Life Technologies). cDNA was generated from total RNA using oligo-dT primers with the SuperScript II RNase H- reverse transcription system (Invitrogen, Carlsbad, CA) according to manufacturer’s instructions. Real-time TaqMan polymerase chain reaction (PCR) was performed according to Applied Biosystems (Foster City, CA) protocols. All samples were loaded in triplicate wells and fluorescence emission detected by the ABI Prism 7000 (Applied Biosystems). Primers and probes for NDST-1 and NDST-2 were designed using the Primer Express Software version 1.5 (Applied Biosystems) as follows:

NDST-1 5′ GGA GGG CAC ACG CAT GA 3′ (sense)

5′ TGT GCG CGT AGT TCG TTC TG 3′ (antisense)

5′ FAM-CAA ACA GGG CCT TCA CGT CCT CCA-TAMRA 3′ (probe)

NDST-2 5′ CAC GGA CCC TCA GGT TTG AT 3′ (sense)

5′ GGC CTA GAC AGC GAG TCT TAC C 3′ (antisense)

5′ FAM-CTT CAA GTC CCT GGC ACC AAA ATC CCT TA 3′ (probe)

Plasmids containing the full-length NDST-1 and NDST-2 cDNA sequences were used for positive control. All results were normalized to the expression of endogenous glyceraldehyde-3-phosphate dehydrogenase mRNA and expressed as a relative increase or decrease in expression.

Statistical Analysis

All experimental groups were compared with the corresponding control using the Student two-sample t test; P values less than 0.05 were considered significant. Data points are presented as mean values ± standard error of mean.

RESULTS

Localization of Glycosaminoglycan and Chemokine Expression During Renal Allograft Rejection

Immunofluorescence LSCM allowed precise localization and semiquantitative comparison of the abundance of HS, C4S, and C6S within normal renal tissue and a range of sections taken from rejection biopsies. Figure 1 shows the expression of HS (Fig. 1A), C4S (Fig. 1B), and C6S (Fig. 1C) in representative sections from a kidney showing acute cellular rejection. Although HS is preferentially expressed at high levels within the basement membrane, both forms of chondroitin sulphate are distributed within the interstitial tissues with only low-level expression associated with the tubules. Normal renal tissue expressed each GAG at a similar location, but the expression of each species was greater during acute rejection (Fig. 2A; n=20, HS P<0.01, C4S P<0.05, C6S P<0.01). Specific comparison of the abundance of N-sulphated HS associated with blood vessel walls (peritubular capillaries) also showed a significant increase during acute rejection (Fig. 2B; n=7, P<0.0001). There was no obvious relationship between the GAG expression and the severity of acute rejection.

F1-12
FIGURE 1.:
Glycosaminoglycan (GAG) expression within renal biopsy sections. Representative results from immunohistochemical investigation of human renal sections. Acutely rejecting renal biopsies were stained with antibodies specific for heparan sulphate (HS) (A), chondroitin-4-sulphate (C4S) (B), and chondroitin-6-sulphate (C6S) (C), respectively. Expression was visualized by staining with a fluorescein isothiocyanate (FITC)-conjugated secondary antibody followed by scanning laser confocal microscopy.
F2-12
FIGURE 2.:
GAG expression in normal and rejecting renal biopsies. (A) Average pixel intensity of fluorescence associated with the expression of HS, C4S, and C6S by rejecting renal tissue (n=20) was compared with values for normal kidney sections (n=12). (B) Expression of HS by defined vascular endothelium was compared between rejecting (n=7) and normal (n=8) kidney sections; a minimum of three vessels were examined per section.

Examination of the Coexpression of Glycosaminoglycan and CCL5

Two-color immunofluorescence was used to assess the relationship between the expression of HS and CCL5 during acute rejection. HS was largely restricted to the tubules with the basement membrane showing the highest expression (Fig. 3A). The chemokine CCL5 had a widespread distribution within the tubular epithelium during acute rejection, with the highest levels being observed on the basolateral cell surface and basement membrane (Fig. 3B). CCL5 was also distributed diffusely within the interstitial matrix, with staining associated with endothelial cells of the peritubular venules (Fig. 3B); some graft-infiltrating leukocytes also showed high-intensity fluorescence for CCL5. Two-color analysis showed that CCL5 was predominantly associated with areas rich in HS expression during acute rejection (Fig. 3C). Analysis of multiple sites allowed the demonstration of a positive correlation between the levels of HS and CCL5 expression (Fig. 3D; r2=0.92, P<0.0001).

F3-12
FIGURE 3.:
Colocalization of HS and chemokine expression during rejection. Dual-color immunofluorescence-labeled sections were analyzed by laser scanning confocal microscopy (LSCM) to visualize HS on the FITC (A) channel and CCL5 on the TRITC (B) channel in a representative renal allograft biopsy with acute rejection; colocalization produces a yellow color (C). (D) Positive correlation between the point fluorescence intensity for HS (green fluorescence per pixel) and CCL5 (red fluorescence per pixel).

Examination of N-Deacetylase/N-Sulphotranferase-1 Upregulation in Primary Human Endothelial Cells

To identify the molecular mediator involved in increasing the chemokine binding to N-sulphated endothelial HS during inflammation, changes in the expression of NDST-1 and NDST-2 were analyzed by real-time PCR using parallel measurement of glyceraldehyde-3-phosphate dehydrogenase for internal control. Primary cultured HUVECs were stimulated with the proinflammatory cytokines IFNγ and TNFα for 24 hr before the analysis of mRNA levels. Analysis of two separate lines of cells showed that the expression of NDST-1 in both was increased (P<0.01) by a mean of sevenfold ± fourfold by stimulation for this time. The expression of NDST-2 showed no increase.

Glycosaminoglycan Expression by Endothelial Cells (Human Microvascular Endothelial Cell Line)

The immortalized human microvascular endothelial cell line HMEC-1 (21) provides several advantages for investigation of aspects of the immunobiology of microvascular endothelium (25), including a reproducible response to proinflammatory cytokines (26). Cultured HMEC-1 cells showed cobblestone morphology and uniformly expressed the endothelial marker CD31 (Fig. 4A). Confocal analysis of HMEC-1 cells revealed bright immunofluorescence after primary HS labeling and appropriate secondary antibody development (Fig. 4B). The primary monoclonal antibody 10E4 was used to define the abundance of N-sulphated epitopes within HS on the endothelial surface (22). Less-intense signals were produced by labeling with anti-C4S (Fig. 4C) and anti-C6S (Fig. 4D) antibodies. In all cases, cells incubated as appropriate with nonimmune mouse IgM or IgG in place of primary antibody produced no significant control fluorescence above that of unlabeled cells. Treatment of HMEC-1 cells with heparitinase reduced subsequent antibody labeling of HS to control levels (results not shown).

F4-12
FIGURE 4.:
GAG expression by cultured endothelial cells. Cells were stained with antibodies specific for HS (A), C4S (B), C6S (C), and CD31 (D), respectively. Relative expression was visualized after secondary labeling with an FITC-conjugated secondary antibody by scanning laser confocal microscopy. Results are representative of five independent experiments.

Examination of the Binding of Biotinylated CCL5 to Human Microvascular Endothelial Cells

Previous studies have shown that sequestration of chemokines by cell-surface GAGs can increase the biological activity of these factors and is essential for chemokine-mediated leukocyte migration (12, 27). By using the prototypical CC chemokine CCL5, which is known to bind the sulphated domains on HS, a series of experiments was performed to determine whether pretreatment of endothelial cells with proinflammatory cytokines can increase the potential for CCL5 sequestration by HS molecules. Monolayers of TNFα- and IFNγ-activated HMEC-1 cells were used to model inflamed vascular endothelium (28). The results of binding biotinylated CCL5 to resting HMEC-1 cells are shown in Figure 5A; the image shows that only a small amount of the exogenous CCL5 was bound by these cells. With similar settings for the confocal microscope, the apparent binding of exogenous CCL5 to HMEC-1 (which had been prestimulated with TNFα and IFNγ) for 24 hr was greatly increased (Fig. 5B); stimulation of the cells with either cytokine alone produced no increase in chemokine binding. The binding of biotinylated CCL5 to cytokine-activated HMEC-1 was decreased when labeling was performed after pretreatment of the cells with heparitinase (Fig. 5C). Quantification of the fluorescence signal by image analysis (Fig. 5D) showed that the mean pixel intensity associated with CCL5-binding was increased twofold after cytokine treatment of the endothelial cells (P<0.001), whereas heparitinase pretreatment reduced chemokine binding to a level significantly below that of resting cells (P<0.001). Application of FITC-avidin in the absence of biotinylated CCL5 produced no nonspecific fluorochrome binding.

F5-12
FIGURE 5.:
Binding of exogenous (biotinylated) CCL5 to cultured endothelial cells. Scanning laser confocal microscopy was used to analyze the binding of biotinylated CCL5 to the apical surface of human microvascular endothelial (HMEC)-1 cells that had been cultured using a range of experimental conditions. Bound CCL5 was detected using a fluorescein/avidin conjugate. CCL5 binding to the surface of resting HMEC-1 (A), CCL5 binding on the surface of HMEC-1 cells that had been stimulated with tumor necrosis factor (TNF)-α and interferon (IFN)γ for 24 hr (B), CCL5 binding to the surface of HMEC-1, which had been prestimulated with TNFα and IFNγ and subsequently treated with heparitinase (C), and mean fluorescence intensity of bound CCL5 (D); mean ± standard error (bars) of mean fluorescence for at least eight cells per specimen from one of three similar experiments.

Examination of the Binding of Biotinylated CCL5 to Renal Sections

To further investigate the chemokine binding potential of GAGs during rejection, sections of rejecting and normal kidney were incubated with biotinylated RANTES. Bound RANTES was visualized by Vector red. In normal kidney there was no significant binding of RANTES (Fig. 6A), whereas sections from acute rejection biopsy specimens showed discrete regions of focal binding of exogenous RANTES within extracellular matrix associated with the tubular basement membrane (Fig. 6B).

F6-12
FIGURE 6.:
Binding of exogenous CCL5 to renal biopsy sections. Cryostat sections from normal human kidney (A) and acute rejection biopsy (B) were incubated with biotinylated regulated on activation normal T-cell expressed and secreted (RANTES). Binding was visualized with Vector red substrate, and sections were analyzed with a Leica DMR microscope.

DISCUSSION

Cytokines including IFNγ, interleukin-2, TGF-β, and the chemokines are known to be sequestered adjacent to their site of synthesis by specific interaction with GAG components of local cell membranes and extracellular matrix. This interaction between cytokine and GAG molecule is known to play a crucial role in creating and stabilizing the microenvironment within which the immune system functions during allograft rejection (reviewed by Ali et al.) (4). Several elegant studies have shown that the level of mRNA encoding individual cytokines can vary with time within a transplanted organ, suggesting similar variation in the expression of cytokine proteins. However, it is now clear that the biological activity of critical inflammatory cytokines, such as the chemokines, is regulated not only by production but also by abundance and presentation in complex with a GAG species. Indeed, GAG nonbinding mutant chemokines have been shown to lack many of their normal biologic activities both in vitro (11, 12) and in vivo (13). Previous studies by our group (19) and others (20) have suggested that both the abundance and biochemistry of GAG species are modified in vivo during periods of inflammation, leading to the possibility that this process might also regulate the local inflammatory response.

To explore the relationship between GAG expression and the sequestration of chemokines in vivo, a series of normal kidney and renal allograft biopsy sections were examined. This tissue was chosen because previous studies have shown that the chemokine CCL5, which has a high affinity for HS (7), can be located precisely and shows strong expression during the inflammation associated with acute rejection (3, 29, 30). Immunofluorescence labeling suggested that N-sulphated HS epitopes were more abundant than either C4S or C6S within the renal sections. HS was predominantly localized within basolateral aspects of the renal tubules and was also concentrated in the tubular basement membrane, where it is known that HS is a major component of the normal tubular basement membrane (31). In contrast, both C4S and C6S were restricted mainly to the interstitial tissues. During acute rejection the expression of each GAG was increased, but HS showed the greatest increase with a tendency for expression within the interstitium and blood vessel walls and the tubular basement membrane. Increased production of hyaluronan in acute and chronically rejecting kidneys has already been demonstrated and also correlates with edema and interstitial inflammation (32).

The chemokine CCL5 was mainly observed within the tubular basement membrane during acute rejection, with less expression within the interstitium except for localized high-level expression associated with individual infiltrating immune cells; this distribution is consistent with previous studies (3, 33). A positive correlation was observed between localized levels of expression of N-sulphated HS and CCL5, suggesting a biological role for HS in the localization of CCL5 and, thus, in targeting leukocyte infiltration to specific sites during rejection. Indeed, it is likely that the distribution observed for CCL5 within rejecting renal tissue defines the sites of active sequestration rather than specific sites of chemokine production, although it is known that activated tubular epithelial cells can produce this chemokine (34).

To model the changes in GAG biology observed during rejection, a series of experiments was performed to examine the potential for treatment of primary vascular endothelial cells with proinflammatory cytokines to up-regulate the expression of mRNA encoding NDST enzymes. It was found that the predominant endothelial isoform of this enzyme, NDST-1, was up-regulated significantly after 24 hr, but that the expression of NDST-2 did not increase. Our group previously correlated similar observations for NDST-1 expression by an immortalized human endothelial cell line with increased expression of N-sulphated epitopes within membrane-associated HS (19). This is consistent with a recent study that demonstrated that cells transfected to express high levels of NDST-1 produced an over-sulphated form of HS (35). Further reports have shown that proinflammatory cytokine treatment also produces an increase in levels of the GAG species HS, C4S, and dermatan sulphate in cell-associated and supernatant fractions from endothelial cells (20), whereas changes in NDST-1 mRNA levels have been observed during experimental nephrosis (36).

In this study it was shown that the apical surface of resting endothelial cells bound little CCL5. However, the potential to bind exogenous CCL5 was increased after stimulation of the cells with the proinflammatory cytokines IFNγ and TNFα, which showed a powerful synergy because neither cytokine alone caused an alteration in chemokine binding. Demonstration that this increase could be blocked by treatment of the endothelial cells with heparitinase, which removes HS from the cell surface, highlights the importance of changes in the expression of this GAG species for the localization of rejection-critical chemokines such as CCL5. This observation is consistent with previous studies that have shown that the HMEC-1 endothelial cell line uses HS to bind a range of exogenous proteins (37). The binding of CCL5 to HS is known to involve ionic interaction between a positively charged BBXB amino acid sequence domain, where B denotes basic amino acid residues, and anionic O- and N-sulphated regions within HS (7). The modulation of CCL5 binding by treatment with heparitinase shows that this chemokine is sequestered solely by endothelial HS and not by any specific chemokine receptor; indeed, previous studies have shown that endothelial cells do not express CCR1 or CCR5, the predominant receptors for CCL5.

Treatment of renal biopsy sections with traceable, biotinylated CCL5 allowed the definition of sites capable of sequestering this chemokine. It was found that normal renal sections bound little CCL5 but that focal sites within the basement membrane acquired the potential to bind increased quantities of this chemokine during acute rejection. The failure to detect significant sequestration of this chemokine by vascular endothelium in these preparations might be a reflection of a lack of sensitivity of this direct labeling technique, suggesting possible benefits of autoradiographic methods that have been used to detect chemokines bound in small quantities by specific receptors (38). However, the results from examination of the tubular basement membrane are entirely consistent with chemokine sequestration by the increased abundance of heavily sulphated HS within this matrix during acute rejection.

This study provides the first evidence that changes in the disposition and composition of GAG species in general, and HS in particular, might contribute to the development of pathology during episodes of acute renal allograft rejection. We specifically showed that the abundance of N-sulphated domains in HS increase within the tubular basement membrane and blood vessel walls during acute rejection. This increase correlates with a localized increase in expression of the chemokine CCL5. Further data suggest that the local increase in the abundance of N-sulphated HS during inflammation is associated with a differential increase in expression of the enzyme NDST-1. Significantly, both the endothelial cell surface in vitro and the tubular basement membrane in vivo show an enhanced capacity to bind CCL5 during inflammation. Because leukocytes respond to immobilized chemokines, it is likely that the increased capacity to present CCL5 at these sites plays a role in recruiting leukocytes from the blood and then in directing responsive cells to the tubules, which are primary targets for damage during acute rejection. It is possible that the critical interaction between chemokines and sulphated GAG species provides a potential site for anti-inflammatory intervention.

ACKNOWLEDGMENTS

The authors thank Dr. Trevor Booth for his help with LSCM and Dr. Noel Carter for helpful discussions.

REFERENCES

1.Robertson H, Kirby JA. Post-transplant renal tubulitis: the recruitment, differentiation and persistence of intra-epithelial T cells. Am J Transplant 2003; 3(1): 3.
2.Hancock WW, Wang L, Ye Q, et al. Chemokines and their receptors as markers of allograft rejection and targets for immunosuppression. Curr Opin Immunol 2003; 15(5): 479.
3.Robertson H, Morley AR, Talbot D, et al. Renal allograft rejection: beta-chemokine involvement in the development of tubulitis. Transplantation 2000; 69(4): 684.
4.Ali S, Hardy LA, Kirby JA. Transplant immunobiology: a crucial role for heparan sulfate glycosaminoglycans? Transplantation 2003; 75(11): 1773.
5.Kuschert GS, Coulin F, Power CA, et al. Glycosaminoglycans interact selectively with chemokines and modulate receptor binding and cellular responses. Biochemistry 1999; 38(39): 12959.
6.Hoogewerf AJ, Kuschert GSV, Proudfoot AEI, et al. Glycosaminoglycans mediate cell surface oligomerization of chemokines. Biochemistry 1997; 36(44): 13570.
7.Lortat-Jacob H, Grosdidier A, Imberty A. Structural diversity of heparan sulfate binding domains in chemokines. Proc Natl Acad Sci U S A 2002; 99(3): 1229.
8.Adams DH, Lloyd AR. Chemokines: leucocyte recruitment and activation cytokines. Lancet 1997; 349(9050): 490.
9.Cinamon G, Shinder V, Alon R. Shear forces promote lymphocyte migration across vascular endothelium bearing apical chemokines. Nat Immunol 2001; 2(6): 515.
10.Hardy LA, Booth TA, Lau EK, et al. Examination of MCP-1 (CCL2) partitioning and presentation during transendothelial leukocyte migration. Lab Invest 2004; 84(1): 81.
11.Proudfoot AE, Fritchley S, Borlat F, et al. The BBXB motif of RANTES is the principal site for heparin binding and controls receptor selectivity. J Biol Chem 2001; 276(14): 10620.
12.Ali S, Fritchley SJ, Chaffey BT, et al. Contribution of the putative heparan sulfate-binding motif BBXB of RANTES to transendothelial migration. Glycobiology 2002; 12(9): 535.
13.Proudfoot AE, Handel TM, Johnson Z, et al. Glycosaminoglycan binding and oligomerization are essential for the in vivo activity of certain chemokines. Proc Natl Acad Sci U S A 2003; 100(4): 1885.
14.Robertson H, Wheeler J, Morley AR, et al. Beta-chemokine expression and distribution in paraffin-embedded transplant renal biopsy sections: analysis by scanning laser confocal microscopy. Histochem Cell Biol 1998; 110(2): 207.
15.Sasisekharan R, Venkataraman G. Heparin and heparan sulfate: biosynthesis, structure and function. Curr Opin Chem Biol 2000; 4(6): 626.
16.Unger E, Pettersson I, Eriksson UJ, et al. Decreased activity of the heparan sulfate-modifying enzyme glucosaminyl N-deacetylase in hepatocytes from streptozotocin-diabetic rats. J Biol Chem 1991; 266(14): 8671.
17.Li J, Hagner-McWhirter A, Kjellen L, et al. Biosynthesis of heparin/heparan sulfate. cDNA cloning and expression of D-glucuronyl C5-epimerase from bovine lung. J Biol Chem 1997; 272(44): 28158.
18.Aikawa J, Grobe K, Tsujimoto M, et al. Multiple isozymes of heparan sulfate/heparin GlcNAc N-deacetylase/GlcN N-sulfotransferase. Structure and activity of the fourth member, NDST4. J Biol Chem 2001; 276(8): 5876.
19.Carter NM, Ali S, Kirby JA. Endothelial inflammation: the role of differential expression of N-deacetylase/N-sulphotransferase enzymes in alteration of the immunological properties of heparan sulphate. J Cell Sci 2003; 116 (Pt 17): 3591.
20.Klein NJ, Shennan GI, Heyderman RS, et al. Alteration in glycosaminoglycan metabolism and surface charge on human umbilical vein endothelial cells induced by cytokines, endotoxin and neutrophils. J Cell Sci 1992; 102 (Pt 4): 821.
21.Ades EW, Candal FJ, Swerlick RA, et al. HMEC-1: establishment of an immortalized human microvascular endothelial cell line. J Invest Dermatol 1992; 99(6): 683.
22.David G, Bai XM, Van der Schueren B, et al. Developmental changes in heparan sulfate expression: in situ detection with mAbs. J Cell Biol 1992; 119: 961.
23.Racusen LC, Solez K, Colvin RB, et al. The Banff 97 working classification of renal allograft pathology. Kidney Int 1999; 55(2): 713.
24.Robertson H, Wong WK, Talbot D, et al. Tubulitis after renal transplantation: demonstration of an association between CD103+ T cells, transforming growth factor beta1 expression and rejection grade. Transplantation 2001; 71(2): 306.
25.Bosse D, George V, Candal FJ, et al Antigen presentation by a continuous human microvascular endothelial cell line, HMEC-1, to human T cells. Pathobiology 1993; 61 (3–4): 236.
26.Unger RE, Krump-Konvalinkova V, Peters K, et al. In vitro expression of the endothelial phenotype: Comparative study of primary isolated cells and cell lines, including the novel cell line HPMEC-ST1.6R. Microvascular Res 2002; 64(3): 384.
27.Ali S, Palmer AC, Banerjee B, et al. Examination of the function of RANTES, MIP-1alpha, and MIP-1beta following interaction with heparin-like glycosaminoglycans. J Biol Chem 2000; 275(16): 11721.
28.Luscinskas FW, Ding H, Tan P, et al. L- and P-selectins, but not CD49d (VLA-4) integrins, mediate monocyte initial attachment to TNF-alpha-activated vascular endothelium under flow in vitro. J Immunol 1996; 157(1): 326.
29.Inston NG, Cockwell P. The evolving role of chemokines and their receptors in acute allograft rejection. Nephrol Dial Transplant 2002; 17(8): 1374.
30.El-Sawy T, Fahmy N, Fairchild R. Chemokines: directing leukocyte infiltration into allografts. Curr Opin Immunol 2002; 14(5): 562.
31.Miner JH. Renal basement membrane components. Kidney Int 1999; 56(6): 2016.
32.Wells A, Larsson E, Hanas E, et al. Increased hyaluronan in acutely rejecting human kidney grafts. Transplantation 1993; 55(6): 1346.
33.Segerer S, Cui Y, Eitner F, et al. Expression of chemokines and chemokine receptors during renal allograft transplant rejection. Am J Kidney Dis 2001; 37: 518.
34.Baer PC, Scherberich JE, Bereiter-Hahn J, et al. Induction of RANTES, HLA-DR, and intercellular adhesion molecule-1 on highly purified distal tubular cells from human kidney. Transplantation 2000; 69(11): 2456.
35.Bengtsson J, Eriksson I, Kjellen L. Distinct effects on heparan sulfate structure by different active site mutations in NDST-1. Biochemistry 2003; 42(7): 2110.
36.Nakayama K, Natori Y, Sato T, et al. Altered expression of NDST-1 messenger RNA in puromycin aminonucleoside nephrosis. J Lab Clin Med 2004; 143(2): 106.
37.Robinson MJ, Tessier P, Poulsom R, et al. The S100 family heterodimer, MRP-8/14, binds with high affinity to heparin and heparan sulfate glycosaminoglycans on endothelial cells. J Biol Chem 2002; 277(5): 3658.
38.Patterson AM, Siddall H, Chamberlain G, et al. Expression of the Duffy antigen/receptor for chemokines (DARC) by the inflamed synovial endothelium. J Pathol 2002; 197(1): 108.
Keywords:

Chemokine; Glycosaminoglycan; Allograft rejection; Chemotaxis; Heparan sulphate

© 2005 Lippincott Williams & Wilkins, Inc.