Impaired Angiogenesis in the Remnant Kidney Model: I. Potential Role of Vascular Endothelial Growth Factor and Thrombospondin-1 : Journal of the American Society of Nephrology

Journal Logo

Pathophysiology of Renal Disease

Impaired Angiogenesis in the Remnant Kidney Model

I. Potential Role of Vascular Endothelial Growth Factor and Thrombospondin-1


Author Information
Journal of the American Society of Nephrology 12(7):p 1434-1447, July 2001. | DOI: 10.1681/ASN.V1271434
  • Free


Regardless of the initial cause, end-stage renal disease is characterized by progressive scarring of the glomeruli (glomerulosclerosis) and the interstitium (interstitial fibrosis). Numerous studies have focused on the pathogenesis of the renal scarring process, including the role of infiltrating leukocytes, vasoactive mediators, cytokines, apoptosis, and extracellular matrix proteins and proteases. To date, most studies on the pathogenesis of progressive renal disease have focused on podocytes and mesangial cells in the glomeruli and tubular cells and fibroblasts in the interstitium, whereas the role of the microvasculature in progressive renal injury and scarring has not been studied in depth.

There is increasing evidence, however, that the microvasculature may play a critically important role in progressive renal disease. Bohle et al. (1) observed a remarkable loss of peritubular capillaries in interstitial fibrosis in human disease, and a loss of peritubular capillaries was also observed in several experimental models of interstitial fibrosis (2,3). A loss of capillaries would result in impaired delivery of oxygen and nutrients to the tubules and interstitial cells, producing chronic ischemia. A particularly vulnerable area would be the outer medulla, which normally exists in a borderline hypoxic environment because of countercurrent circulation and the high metabolic demands of the tubular epithelial cells in the thick ascending limb (4). Hypoxia has been demonstrated to induce tubular cell and fibroblast proliferation, matrix synthesis, cytokine release, and upregulation of tubular cell Fas expression (5,6,7). Fibrotic scarring resulting from ischemia would exacerbate hypoxia by increasing the oxygen diffusion gradient; therefore, chronic hypoxia may have a crucial role in progressive renal disease.

Yamanaka and co-workers (8,9) also observed a loss of glomerular endothelial cells in two experimental models of progressive glomerulosclerosis. In both the remnant kidney (RK) model and anti-glomerular basement membrane (anti-GBM) disease, an ineffectual glomerular endothelial proliferative response was demonstrated in association with progressive endothelial cell loss. Those authors postulated that glomerular endothelial loss results in denudation of the GBM, with activation of the coagulation system, capillary collapse, and subsequent glomerulosclerosis (8,9).

The pathogenesis of the progressive capillary loss in chronic renal disease has not been determined. Although there are likely to be specific factors that mediate endothelial cell death, such as oxidants and angiotensin II (10,11), there may also be ineffectual repair (2,3,8,9). This contrasts with acute glomerular injury, in which complete capillary repair is common (12,13). In this study, we examined potential mechanisms for the inadequate repair of the endothelium in the RK model, to investigate the role of the renal microvasculature in progressive renal disease.

Materials and Methods

Experimental Protocol

All animal procedures were conducted after approval of the protocol by the University of Washington Animal Care Committee. Male Sprague-Dawley rats (200 to 240 g) underwent baseline BP and renal function assessments and were randomly assigned to the RK group or the sham-operated control group. For the RK group (n = 24), a right subcapsular nephrectomy was performed and the right kidney was immediately fixed for baseline histologic evaluation. This procedure was followed by surgical resection of the upper and lower thirds of the left kidney. The renal mass reduction was performed by surgical resection instead of ligation of the renal arterial branches because the former method results in chronic renal failure with minimal elevations in BP (14) and thus minimizes the effect of systemic hypertension on the renal microvasculature. Strict hemostasis and aseptic techniques were used during the surgical procedure. For the control group (n = 16), a sham operation, consisting of a laparotomy and manipulation of the renal pedicles but without the destruction of renal tissue, was performed. All animals were fed a standard laboratory diet and water ad libitum.

Preoperative and postoperative 24-h urinary protein excretion rates were measured using the sulfosalicylic acid method, and blood urea nitrogen (BUN) levels were determined colorimetrically with a commercial kit (Sigma Diagnostics, St. Louis, MO). Systolic arterial BP was monitored with a tail cuff sphygmomanometer, using an automated system with a photoelectric sensor (IITC; Life Sciences, Woodland Hills, CA), which has been demonstrated to be closely correlated with intra-arterial measurement of BP (15).

Renal Morphologic and Immunohistochemical Analyses

Tissue for light microscopy and immunoperoxidase staining was fixed in methyl Carnoy's solution and embedded in paraffin. Four-micrometer sections were stained with the periodic acid-Schiff re-agent and counterstained with hematoxylin. Indirect immunoperoxidase staining of 4-μm sections was performed as described previously (16), with the following specific monoclonal and polyclonal antibodies: endothelial cells were detected with the mouse monoclonal antibody JG-12, directed against a 70-kD cell membrane antigen present on rat endothelial cells (16), and the monoclonal anti-endothelial cell antibody RECA-1 (a gift of Adrian Duijvestijn; University of Limberg, The Netherlands) (17); vascular endothelial growth factor (VEGF) with a rabbit polyclonal antibody (Santa Cruz Biotechnology, Santa Cruz, CA); thrombospondin-1 (TSP-1) with the mouse monoclonal antibody A6.1 (Neomarkers, Fremont, CA); and monocytes/macrophages with the mouse monoclonal antibody ED-1 (Serotec, Indianapolis, IN). Control experiments included omission of the primary antibody and substitution of the primary antibody with preimmune rabbit or mouse serum.

To examine whether there was any evidence of endothelial cell proliferation, double-immunostaining was performed with an anti-endothelial cell antibody (JG-12, and IgG antibody) and an antibody to the proliferating cell nuclear antigen (PCNA) (19A2, an IgM monoclonal antibody; Coulter, Hialeah, FL). The anti-PCNA antibody and JG-12 were incubated simultaneously overnight at 4°C, followed sequentially by incubation with biotinylated rabbit anti-mouse IgM serum, incubation with peroxidase-conjugated avidin D (Vector, Burlingame, CA) color development with diaminobenzidine (DAB) and nickel chloride, and incubation in 3% H2O2 for 8 min, to eliminate any remaining peroxidase activity. Subsequently, sections were incubated with biotinylated horse anti-mouse IgG for 30 min at room temperature, followed by peroxidase-conjugated avidin D and DAB (16).

To examine the relationship between VEGF expression and macrophage infiltration, double-immunostaining with the anti-VEGF antibody (a rabbit polyclonal antibody) and an antibody to a cell-specific marker for macrophages (ED-1, a mouse monoclonal IgG1 antibody) was performed using an indirect immunoperoxidase technique. Samples were incubated with ED-1 overnight at 4°C, followed sequentially by biotinylated rabbit anti-mouse IgG1 (Zymed, San Francisco, CA) and peroxidase-conjugated avidin D, with color development using DAB and nickel chloride. After incubation in 3% H2O2 for 8 min, the anti-VEGF antibody was applied for 2 h at room temperature, followed by biotinylated rat anti-rabbit IgG antibody (Vector) for 30 min and peroxidase-conjugated avidin D and DAB without nickel chloride. Control experiments included omission of either the primary or secondary antibody.

Quantification of Morphologic Data

The number of glomerular capillary loops per glomerular cross-section, as identified by positive JG-12 staining, was counted by a single observer in all glomeruli of tissue sections, at × 400 magnification. Glomerular capillary density was also measured and defined as the number of capillaries per glomerular cross-sectional area (per 0.01 mm2); the latter was determined by morphometric measurements using computer image analysis (Optimas 6.2; Media Cybernetics, Silver Springs, MD). Peritubular capillary densities were quantified in two ways. The area of peritubular capillary staining by JG-12 was expressed as the percent positive area per 100 cortical tubules using computer image analysis, to account for changes in tubular size and to exclude the effects of tubular dilation and/or atrophy on peritubular capillary numbers. The other index, i.e., peritubular capillary rarefaction index, was determined by counting the numbers of squares in 10 × 10 grids that did not contain JG-12-positive peritubular capillary staining, in at least 20 nonoverlapping sequential fields, at ×100 magnification. The minimal possible capillary rarefaction index is 0, i.e., every square in the grid contains a JG-12-positive peritubular capillary, whereas the maximal score is 100, i.e., JG-12-positive peritubular capillaries are absent from every square in the grid (18). Changes in capillary density were confirmed by staining tissue sections with RECA-1, which is an antibody to a different endothelial cell antigen. The mean numbers of proliferating endothelial cells (JG-12-and PCNA-positive cells) in glomeruli and peritubular areas in each biopsy were calculated in a blinded manner, as the mean numbers of positive cells in individual glomeruli and 0.25-mm2 grids, respectively, at ×200 magnification.

The percentage areas of the cortex and outer medulla that were positive for VEGF were measured by computer image analysis (Optimas 6.2) (16). In each biopsy, the negative background staining was calibrated to zero, and the area of positive staining above the background level in each field was measured. Each measurement was derived from computer analysis of the integrated logarithm of the inverse gray value, which is proportional to the total amount of absorbing material in the light path. This system enables the percentage area of positive staining in each biopsy to be accurately quantified. Glomerular and periglomerular TSP-1 expression was graded by counting the percentage of positive glomeruli or glomeruli surrounded by TSP-1 staining, respectively. Tubulointerstitial TSP-1 expression was quantified by counting the number of tubules exhibiting positive TSP-1 staining in 10 sequentially selected, 1-mm2 grids, at ×100 magnification. The mean numbers of macrophages (ED-1-positive cells) in glomeruli and the interstitial area in each biopsy were calculated, in a blinded manner, by averaging the total number of positive cells in each glomerulus or in 30 sequentially selected 0.25-mm2 grids, respectively, at ×200 magnification.

The percentage of glomeruli exhibiting focal or global glomerulosclerosis was determined by evaluation of all glomeruli present in the biopsy. Glomerulosclerosis was defined as segmental increases in the glomerular matrix, segmental collapse, obliteration of capillary lumina, and accumulation of hyaline, often with synechial attachment to Bowman's capsule. Tubulointerstitial injury was defined as inflammatory cell infiltration, tubular dilation and/or atrophy, or interstitial fibrosis. Injuries were graded semiquantitatively by a blinded observer, who examined at least 40 cortical fields (magnification, ×100) in periodic acid-Schiff-stained biopsies. Only cortical tubules were included in the following scoring system (16): 0, normal; 1, involvement of <10% of the cortex; 2, involvement of 10 to 25% of the cortex; 3, involvement of 26 to 50% of the cortex; 4, involvement of 51 to 75% of the cortex; 5, extensive damage involving >75% of the cortex.

Western Blot Analyses

Isolation of whole protein from the RK was performed by tissue homogenization in Tris-glycine buffer with proteinase inhibitors (Complete; Roche, Indianapolis, IN). After determination of the protein concentration using the Bio-Rad protein assay (Bio-Rad, Richmond, CA), protein samples (30 μg) were mixed with reducing buffer, boiled, resolved on 7.5% sodium dodecyl sulfate (SDS)-polyacrylamide gels, and transferred to nitrocellulose membranes by electroblotting. Membranes were blocked with 5% (wt/vol) nonfat milk powder in Tris-buffered saline for 30 min at room temperature. An affinity-purified rabbit polyclonal antibody to human VEGF (Santa Cruz Biotechnology), which recognizes all four isoforms of human VEGF and is reported to cross-react with rat VEGF, was used. After incubation of the membrane with alkaline phosphatase-conjugated mouse anti-rabbit antibody, the bands were observed using 5-bromo-4-chloro-3-indolyl phosphate/nitro blue tetrazolium tablets (Sigma Chemical Co., St. Louis, MO). Positive immunoreactive bands were quantified by densitometry.

In Vitro Effects of Macrophage-Derived Cytokines on Tubular VEGF Expression

Murine thick ascending limb (mTAL) cells were obtained by microdissection of normal mouse (C57BL/6J) kidneys. Cells were identified as pure mTAL cells by their uniform cobblestone appearance when grown to confluence and by their uniform positive staining for Tamm-Horsfall protein, Na+/K+-ATPase, and inducible nitric oxide synthase. For passage of mTAL cells, confluent cells were washed with Hanks' balanced salt solution (HBSS) (Irvine Scientific, Santa Ana, CA) and then incubated for 5 min at 37°C with 1 ml of HBSS containing 0.05% trypsin (Irvine Scientific) and 0.01% ethylenediaminetetraacetate (Irvine Scientific), until the cells were detached. Free cells were suspended in mTAL medium containing 14% NuSerum (Becton-Dickson, Franklin Lakes, NJ), hydrocortisone (40 ng/ml), thyroxine (2 ng/ml; Sigma), penicillin, and streptomycin (Life Technologies, Grand Island, NY), centrifuged at 1200 rpm for 5 min, and seeded. All data presented are from experiments performed with cells from passages 8 to 18.

When the mTAL cells had grown to 70% confluence in multiwell plates (Becton Dickinson), the 14% NuSerum-containing medium was removed and the cells were washed three times with HBSS. The culture medium was changed to serum-free mTAL medium for 24 h before each experiment. After synchronization of cell growth for 24 h, mTAL cells were washed three times with HBSS and exposed to interleukin-1β (IL-1β) (0.1 to 100 ng/ml), IL-6 (0.1 to 100 ng/ml), or tumor necrosis factor-α (TNF-α) (0.1 to 100 ng/ml) (all from Sigma) for 3 to 24 h. All experiments were also conducted under hypoxic conditions. To achieve hypoxia, mTAL cells, with or without cytokines, were incubated in a GasPak anaerobic culture pouch (BBL Microbiology Systems, Kansas City, MO), using hydrogen and a palladium catalyst to remove all traces of oxygen. At the end of a 24-h exposure to hypoxia, the partial pressure of oxygen in the culture medium was measured using a pH/blood gas analyzer; values were in the range of 25 to 30 mmHg and 120 to 130 mmHg for hypoxic and normoxic conditions, respectively.

To quantify the level of VEGF secretion under the different conditions, VEGF protein was measured in mTAL cell culture supernatants using a commercial mouse enzyme-linked immunosorbent assay kit for VEGF (R & D Systems, Minneapolis, MN), which is sensitive to levels of 3 pg/ml. The interassay coefficient of variation was <7% and the intra-assay coefficient of variation was <5% among the standards. All data for VEGF production determined in enzyme-linked immunosorbent assays were expressed as picograms per 105 cells.

Cell viability under the experimental conditions was examined by using lactate dehydrogenase (LDH) assays. After exposure of cells to various cytokines under normoxic or hypoxic conditions, supernatants were collected and filtered (to remove any dead and detached cells), and the LDH activity was measured using an enzymatic method based on the oxidation of NADH to NAD during the reduction of pyruvate, which is catalyzed by LDH (Sigma). LDH release was calculated as a percentage of LDH amounts in the supernatant, compared with total LDH amounts in the lysed mTAL cells plus the supernatant.

RNA Isolation and Reverse Transcription-PCR of VEGF mRNA

After incubation of cells with various proinflammatory cytokines (IL-1β, IL-6, or TNF-α, 10 ng/ml) under normoxic or hypoxic conditions, total RNA was prepared from the mTAL cell monolayers using the RNeasy96 total RNA isolation protocol (Qiagen, Valencia, CA). Total RNA was similarly isolated from cortical and medullary tissue samples from both sham-operated kidneys and RK at 8 wk (n = 4 each).

Total RNA was analyzed by quantitative real-time PCR (19), using an ABI Prism 7700 sequence detection system (PE Applied Biosystems, Foster City, CA). This system is based on the ability of the 5′-nuclease activity of Taq polymerase to cleave a nonextendable, dual-labeled, fluorogenic hybridization probe during the extension phase of PCR. The probe is labeled with a reporter fluorescent dye 6-carboxyfluorescein) at the 5′ end and a quencher fluorescent dye (6-carboxytetramethyl-rhodamine) at the 3′ end. When the probe is intact, reporter emission is quenched by the physical proximity of the reporter and quencher fluorescence dyes. However, during the extension phase of PCR, the nucleolytic activity of the DNA polymerase cleaves the hybridization probe and releases the reporter dye from the probe, with a concomitant increase in reporter fluorescence.

The following sequence-specific primers and probes for mouse VEGF and 18S rRNA were designed using Primer Express software (PE Applied Biosystems): for VEGF: forward, 5′-GGAGCAGAAAGCCCATGAAGT-3′; reverse, 5′-GTCTCAATCGGACGGCAATAG-3′; probe, 5′-6-carboxyfluorescein-TGAAGTTCATGGACGTCTACCAGCGCA-6-carboxytetramethyl-rhodamine-3′; for 18S rRNA: forward, 5′-CGGCTACCACATCCAAGGAA-3′; reverse, 5′-GCTGGAATTACCGCGGCT-3′; probe, 5′-6-carboxyfluorescein-TGCTGGCACCAGACTTGCCCTC-6-carboxytetramethyl-rhodamine-3′.

Primers were used at a concentration of 200 nM and probes at a concentration of 100 nM in each reaction. Multiscribe reverse transcriptase and AmpliTaq Gold polymerase (PE Applied Biosystems) were used for all reverse transcription (RT)-PCR assays. Relative quantitation of 18S rRNA and VEGF mRNA was performed using the comparative threshold cycle number for each sample, fitted to a five-point standard curve (ABI Prism 7700 User Bulletin 2; PE Applied Biosystems). Expression levels were normalized to 18S rRNA levels and related to relevant control values.

Northern Blot Analysis of VEGF mRNA

Total RNA was also analyzed by Northern blot analysis. After spectrophotometric determination of RNA purity and concentration, 20-μg RNA samples were denatured and subjected to electrophoresis through 1% agarose gels containing 2.2 M formaldehyde. RNA was then transferred to nylon membrane filters (Hybond-N; Amersham, Piscataway, NJ) by capillary action and was cross-linked using a Stratalinker (Stratagene, La Jolla, CA), followed by prehybridization for 30 min at 37°C with ExpressHyb hybridization solution (Clontech, Palo Alto, CA). The probe for VEGF (kindly provided by Dr. Mark E. Cooper, University of Melbourne, Victoria, Australia) was labeled with [32P]dCTP by random-primed DNA synthesis. Hybridization was performed at 37°C for 1 h, and then filters were washed three times for 20 min each [first wash in 2× SSC/0.1% SDS at 65°C, second wash in 1× SSC/0.1% SDS at 65°C, and third wash in 0.1× SSC/0.1% SDS at room temperature] before exposure to x-ray film (Kodak). The relative autoradiographic intensities of the Northern blots were determined by scanning densitometry. All results were corrected for differences in RNA loading by rehybridization with an oligonucleotide probe for 18S rRNA.

Statistical Analyses

All data are presented as mean ± SD. Differences in the various parameters between the sham and RK groups were evaluated by unpaired comparisons for nonparametric data. Differences in parameters at each time point after RK surgery were compared by paired t test. The relationships between variables were assessed by Pearson correlation analysis. Significance was defined as P < 0.05.


Induction of Renal Failure, Proteinuria, and Renal Injury by 5/6 Renal Ablation

As shown in Table 1, body weight at each time point after renal ablation was significantly lower for the RK group, compared with the control group. As expected, an immediate increase in BUN levels was observed for the RK rats. BUN levels then remained stable until week 4, after which an additional increase was noted (Table 1). Pathologic proteinuria (defined as >25 mg/24 h) was present in the RK rats after 2 wk and then exhibited a continuous increase up to week 8. In addition, the RK group exhibited a mild but significant increase in systolic BP at week 8, compared with both the baseline value and values for the control group (Table 1). Significant light microscopic changes indicating focal or diffuse glomerulosclerosis were absent at weeks 1 and 2 after renal ablation and gradually increased thereafter in the RK group (Table 1). Significant evidence of tubulointerstitial injury after the 5/6 nephrectomy was first noted at week 2 and progressively increased thereafter.

Table 1:
General characteristics of the RK modela

Changes in the Renal Microvasculature

The number of JG-12-positive glomerular capillary loops per glomerular cross-section revealed no significant changes during the first 4 wk after renal ablation, although there was a trend toward an increase in capillary number, compared with the baseline value (Figure 1A). However, at 8 wk after renal ablation, the number of glomerular capillary loops per glomerular cross-section was significantly reduced (Figure 1, A, C, and D). Glomerular capillary density expressed as the number of capillary loops per 0.01 mm2 of glomerular cross-sectional area was more markedly decreased at week 8, compared with sham-operated rats (14.2 ± 5.2 versus 30.1 ± 8.4 loops/0.01 mm2, RK versus sham, P < 0.05). Peritubular capillary loss, expressed as an increased capillary rarefaction index (i.e., percentage area with no capillaries) and decreased capillary density (i.e., amount of capillary staining/100 tubules), was observed earlier than the changes in glomerular capillaries. Focal peritubular capillary loss was present by 2 wk after renal ablation (Figure 1B) and further increased with time Figure 1, B, E, and F). At 8 wk, the peritubular capillary rarefaction index was markedly increased up to 22.5% and this increase was associated with a significant decrease in peritubular capillary density (10.2 ± 1.8 versus 0.78 ± 0.23%, baseline versus 8 wk, P < 0.01), indicating significant capillary loss with considerable areas of renal tissue devoid of capillaries.

Figure 1:
Changes in the renal microvasculature in the remnant kidney (RK) model. Shown are the changes in the glomerular capillary loop number (A) and the peritubular capillary rarefaction index (B) in RK rats and sham-operated rats during the 8-wk period after surgery. In contrast to the normal microvasculature present in sham-operated rats (C, RECA-1; E, JG-12), in the RK model there was a loss of both glomerular capillary loops (D, week 8, RECA-1) and peritubular capillaries (F, week 4, JG-12), which was most prominent at the 8-wk time point. Endothelial cells were stained with the antibodies RECA-1 and JG-12 using indirect immunoperoxidase techniques. Data are expressed as mean ± SD. Magnifications: ×630 in C; ×400 in D; ×100 in E; ×50 in F.

PCNA and JG-12 double-staining documented peritubular capillary endothelial cell proliferation beginning at 1 wk, followed by glomerular endothelial cell proliferation at 2 wk after the 5/6 nephrectomy (Figure 2). Both glomerular and peritubular capillary endothelial cell proliferation then subsided and eventually decreased to levels below those observed for the control group, at weeks 8 and 4, respectively (Figure 2).

Figure 2:
Evidence that endothelial cell proliferation occurs early in the RK model. The kinetics of proliferating glomerular and peritubular capillary endothelial cells [defined as proliferating cell nuclear antigen (PCNA)- and JG-12-positive cells] in the RK animals during the 8-wk period after surgery are presented. Initial early proliferation of peritubular (week 1) and glomerular (week 2) endothelial cells was apparent, but there was a subsequent decrease in cell proliferation. At week 8, both glomerular and peritubular capillary proliferation rates were lower than those in sham-operated control animals. Data are expressed as mean ± SD. *, P < 0.05 versus sham.

Similar changes in the microvasculature were noted with an antibody to a different endothelial cell antigen (RECA-1), suggesting that the findings observed were likely attributable to true changes in the microvasculature, as opposed to alterations in endothelial antigen expression.

Renal VEGF Expression

Figure 3 presents the changes in renal VEGF expression detected by immunohistochemical staining. In sham-operated rats, there was constitutive expression of VEGF in tubules within the outer medulla and the medullary rays (Figure 3A), similar to that observed in normal kidneys (20,21). No significant change in the percent positive area of renal VEGF expression (which reflects primarily tubular expression) was noted during the first 2 wk after renal ablation. However, a decrease in VEGF immunostaining became evident 4 wk after the 5/6 nephrectomy, with marked change by week 8 (Figure 3, B and C).

Figure 3:
Tubular vascular endothelial growth factor (VEGF) expression in the RK model. In sham-operated rats, VEGF is constitutively expressed in tubules of the outer medulla and the medullary rays (A, arrows). In rats with RK, there was a marked decrease in tubular VEGF expression (B, week 4). Quantification of VEGF expression by computer image analysis (C) is shown for the RK group and the control group. VEGF was identified by immunostaining (see Materials and Methods section). Data are expressed as mean ± SD. Magnification: ×25 in A and B.

In sham-operated rats, VEGF was constitutively expressed in podocytes of the glomeruli (Figure 4A), as in normal kidneys (20,21). In the RK model, glomerular VEGF immunostaining in podocytes demonstrated a mild increase at weeks 1 and 2. However, VEGF staining in podocytes was markedly decreased by 8 wk (Figure 4, B and C), especially in glomeruli with macrophage infiltration (Figure 4B). Figure 5 presents the results of RT-PCR and Western blot analyses of RK tissue (8 wk) for VEGF, revealing the downregulation of VEGF mRNA and protein in both the cortex and medulla in this model of progressive renal disease (Figure 5).

Figure 4:
Glomerular VEGF expression in the RK model. In sham-operated rats, VEGF is constitutively expressed in podocytes (A, arrowheads). In rats with RK, there was a significant decrease in VEGF immunostaining in podocytes (B and C, week 8, thick arrows), especially in glomeruli with macrophage infiltration. Macrophages (thin arrows) are darker brown than VEGF staining. Glomeruli with greater numbers of ED-1-positive cells exhibited less VEGF staining in podocytes, as assessed by double-immunostaining with ED-1 and anti-VEGF (B), whereas glomerular VEGF expression was still evident within the glomeruli without macrophage infiltration (C). Magnifications: ×630 in A; ×400 in B and C.
Figure 5:
Reverse transcription (RT)-PCR and Western blot analysis in the RK model. RT-PCR (A, n = 4) and Western blot analysis (B) of RK tissue (8 wk) for VEGF demonstrated marked downregulation of VEGF mRNA and protein in both the cortex (C) and medulla (M). VEGF protein expression in the cortex and medulla of the RK rats were reduced by 68 and 62%, respectively, compared with sham-operated rats. Data are expressed as mean ± SD. Shown in B is a representative blot of samples from two sham-operated and two RK rats at 8 wk.

Renal TSP-1 Expression

De novo expression of TSP-1 in glomerular and tubular areas was evident 1 and 2 wk after renal ablation, respectively. TSP-1 was first observed in peritubular and periglomerular areas and in platelets and platelet aggregates in glomeruli (Figure 6, A and B). TSP-1 expression was increased in interstitial areas and glomeruli at 1 and 2 wk, respectively, and the increased expression was sustained at 8 wk (Figure 6, C to F).

Figure 6:
Thrombospondin-1 (TSP-1) expression in the RK model. In sham-operated rats, weak expression of TSP-1 can be observed in rare parietal epithelial cells (A). In rats with RK, TSP-1 was first observed in peritubular areas (B). TSP-1 expression was markedly increased in both glomeruli (C, week 8) and tubules (D, week 4). The amounts of both glomerular (E) and tubular (F) TSP-1 staining increased progressively with time. TSP-1 was identified by immunostaining (see Materials and Methods). Data are expressed as mean ± SD. Magnifications: ×400 in A, B, and C; ×100 in D.

Macrophage Infiltration and Renal Expression of VEGF and TSP-1

There was a significant increase in the number of ED-1-positive cells within both the glomeruli and the tubulointerstitium of the RK group. The increase in the number of ED-1-positive cells began at 1 wk and inexorably increased thereafter (Table 2).

Table 2:
Changes in macrophage infiltration in the RK modela

Double-immunostaining with ED-1 and anti-VEGF revealed that the glomeruli with the most marked macrophage infiltration exhibited more dramatic reductions in VEGF staining in podocytes (Figure 4B). Conversely, glomerular VEGF expression was preserved in glomeruli without significant macrophage infiltration (Figure 4C). Interestingly, tubulointerstitial macrophage infiltration was also spatially and quantitatively associated with a significant decrease in VEGF immunostaining in the cortex and the outer and inner medulla (Figure 7, A through C). For individual animals, a significant inverse correlation between the number of tubulointerstitial macrophages and the percent positive area of VEGF immunostaining was observed (Figure 7C). Similarly, there was a positive correlation between TSP-1 expression and the number of glomerular and interstitial macrophages (Figure 7, D and E). The number of ED-1-positive cells in the tubulointerstitium was also correlated with the loss of peritubular capillaries, as quantified by the peritubular capillary rarefaction index (Figure 7F).

Figure 7:
Relationship of macrophage infiltration with tubular VEGF expression in the RK. There was a dramatic loss of tubular VEGF in areas with macrophage accumulation (A, week 2; B, week 8). (Macrophages are darker brown than VEGF staining.) The sites of macrophage accumulation were correlated quantitatively with both the loss of tubular VEGF (C) and the presence of TSP-1 in tubulointerstitium (D) and the glomeruli (E), as well as with the degree of peritubular capillary rarefaction (F). Double-labeling was performed by staining tissue sections for both macrophages (ED-1-positive cells) and VEGF (see Materials and Methods section). Magnifications: ×50 in A; ×200 in B.

Relationship between Capillary Loss and Changes in VEGF and TSP-1 Expression

There was a significant inverse correlation between tubular TSP-1 expression and peritubular capillary density (r2 = -0.69, P < 0.05), as well as between glomerular TSP-1 expression and glomerular capillary loop number per glomerular cross-section (r2 = -0.55, P < 0.05). Tubular VEGF expression was also correlated with peritubular capillary density (r2 = 0.66, P < 0.01).

Relationship between Changes in the Renal Microvasculature and Renal Structure and Function

The glomerular capillary number per glomerular cross-section and the peritubular capillary density demonstrated significant inverse correlations with 24-h urinary protein excretion (r2 = -0.63 and -0.57, respectively; P < 0.05). A weak but significant correlation between the peritubular capillary rarefaction index and BUN levels was also observed from 2 wk after renal mass reduction (r2 = 0.46, P < 0.05). Glomerular capillary numbers were correlated with the percentage of glomeruli demonstrating sclerosis (r2 = -0.55, P < 0.05). Interstitial fibrosis scores were also inversely correlated with the peritubular capillary density (r2 = -0.80, P < 0.05).

Evidence that Macrophage-Derived Cytokines Decrease VEGF Expression in mTAL Cells

The observation that there was a spatial and quantitative correlation of macrophages with sites of decreased VEGF expression suggested that macrophages could be releasing factors that might modulate VEGF expression. We therefore conducted studies to determine whether several of the more important macrophage proinflammatory cytokines could regulate VEGF expression in renal tubular cells. mTAL cells were exposed to IL-1β, IL-6, or TNF-α (range, 0.1 to 100 ng/ml). Exposure of mTAL cells to IL-1β or IL-6 for 24 h significantly inhibited VEGF protein secretion (Figure 8A). TNF-α decreased VEGF secretion only at the highest dose (100 ng/ml) tested. At the concentrations used, there was no evidence for cytotoxicity, as documented by LDH release. We further investigated the effects of these cytokines in altering VEGF expression in response to hypoxia. Hypoxia increased VEGF protein secretion by 2.4-fold, compared with normoxic control animals (P < 0.05). Similar to their effects on normoxic cells, all three cytokines also inhibited VEGF secretion (Figure 8B). The percentage inhibition of VEGF protein secretion by IL-1β was comparable under normoxic and hypoxic conditions. However, IL-6 and TNF-α induced more profound decreases in VEGF protein secretion under hypoxic conditions.

Figure 8:
Effect of macrophage-derived cytokines on VEGF protein secretion by murine medullary thick ascending limb (mTAL) cells. Stimulation of mTAL cells with interleukin-1β (IL-1β), IL-6, or tumor necrosis factor-α (TNF-α) for 24 h resulted in a significant inhibition of the VEGF protein secretion (A, n = 6). A comparison of the effects of each cytokine (10 ng/ml) under normoxic and hypoxic conditions is also shown (B, n = 6). The cytokine-induced decrease in VEGF protein secretion by mTAL cells was also observed in an hypoxic environment. Data are expressed as mean ± SD (*, P < 0.05 versus control; #, P < 0.05 versus control and 0.1 and 1 ng/ml cytokine).

IL-1β, IL-6, and TNF-α also reduced VEGF mRNA expression in mTAL cells, as measured in real-time RT-PCR and Northern blot analyses. Both types of analyses documented significant reductions in VEGF mRNA expression in mTAL cells in response to IL-1β, IL-6, and TNF-α, under both normoxic and hypoxic conditions (Figure 9).

Figure 9:
Effects of macrophage-derived cytokines on VEGF mRNA expression by mTAL cells. Exposure of mTAL cells to macrophage-associated cytokines for 8 h resulted in reduced VEGF mRNA expression, as detected by both RT-PCR (A, n = 6) and Northern blot analysis (B, n = 4). The experiments were performed under hypoxic conditions. Data are expressed as mean ± SD (*, P < 0.05 versus control). C, control.


We have hypothesized that microvascular injury, with consequent tissue hypoxia and ischemia, may play an important role in progressive renal disease. In this study, we used the RK model (5/6 nephrectomy model), which is a classic model of progressive renal scarring with both glomerulosclerosis and interstitial fibrosis (14). We performed the renal mass reduction by surgical resection of the poles of the kidney (“polectomy”), rather than by renal artery ligation (“infarction” model), to prevent the development of severe hypertension (14), which would be a significant confounding factor in the assessment of the effects of microvascular injury on renal progression.

Our first finding was that an initial angiogenic response follows subtotal renal ablation, as documented by an increased level of proliferation of peritubular and glomerular endothelial cells by weeks 1 and 2, respectively. Yamanaka and co-workers (8) also reported early proliferation of glomerular endothelial cells, and other groups reported proliferation of mesangial cells, tubular cells, and fibroblasts (22,23). Morphometric studies also documented early increases in both the length and number of glomerular capillaries (24,25), suggesting angiogenesis. The mechanism of this increased endothelial cell proliferation is unknown but may be related to mesangial cell production of growth factors such as platelet-derived growth factor and transforming growth factor-β (TGF-β) (26,27), to platelet-associated factors (27), or to shear stress-induced activation of endothelial cells, with the release of TGF-β (26). Although TGF-β and other factors may well contribute to the early angiogenic response (28), we also observed that there was an increase in podocyte expression of the potent angiogenic factor VEGF at weeks 1 and 2 after renal ablation. The observation that VEGF expression increases in podocytes in association with endothelial cell proliferation suggests that VEGF may have a role in this process, perhaps via diffusion across the GBM because of the strong affinity of VEGF for heparan sulfate (29).

Unfortunately, the initial proliferative response by the glomerular and peritubular capillary endothelium was not sustained and there was progressive capillary loss that was significant as early as week 4 for the peritubular capillaries and by week 8 for the glomerular capillaries. Although factors mediating endothelial cell death, such as oxidants, angiotensin II, and Fas (10,11), are likely to be important in the progressive endothelial cell loss, the observation in this study that endothelial cell proliferation subsided to levels significantly below those observed for both sham-operated and normal rats demonstrates that an impaired angiogenic response is also contributory. These data are consistent with recent findings that impaired angiogenesis and/or endothelial cell proliferation occurs in other models of progressive renal disease, such as observed with aging or anti-GBM disease (2,3,9). Kitamura et al. (8) also reported impaired glomerular endothelial proliferation late in the RK model. The inability to stimulate endothelial repair in the setting of endothelial cell loss would be expected to amplify the microvascular disease and contribute to progression.

In this study, we investigated whether the impaired angiogenic response was attributable to loss of an angiogenic factor, increased expression of an antiangiogenic factor, or a combination of both. One of the most important angiogenic factors is VEGF, and this factor is constitutively expressed in podocytes and in tubular epithelial cells of the outer medulla and the medullary rays (20). In addition to being an angiogenic factor, VEGF is an endothelial cell survival factor under a variety of conditions (30,31). We observed a loss of VEGF in both podocytes and the tubules of the outer medulla in association with reduced glomerular and peritubular capillary densities and the development of glomerulosclerosis and interstitial fibrosis. These data suggest that a loss of VEGF could be responsible for the impaired angiogenesis and may contribute to the progressive renal scarring. We have noted a similar loss of VEGF in podocytes and tubules in aging-associated glomerulosclerosis and interstitial fibrosis in rats (32). In addition, a loss of VEGF has been observed in podocytes in glomerulosclerosis (33) and in the outer medullary tubules in chronic interstitial disease in human patients (21), suggesting that downregulation of VEGF expression may be important in human renal disease.

The impaired angiogenic response could also be attributable to increased expression of antiangiogenic factors. TSP-1 is an attractive candidate because it has many actions that oppose those of VEGF, in that it inhibits both basic fibroblast growth factor- and VEGF-mediated endothelial cell proliferation (34) and independently induces endothelial cell apoptosis (35). TSP-1 is also expressed in the interstitium in both experimental (36,37) and human (38) interstitial fibrosis and is a strong predictor of the subsequent development of interstitial fibrosis (37). Although TSP-1 may contribute to the development of fibrosis because of its ability to activate TGF-β (39), our observation that TSP-1 levels were increased in glomeruli and the interstitium and were correlated with the loss of glomerular and peritubular capillaries suggests that TSP-1 may promote renal scarring via effects on the endothelium.

We also addressed potential mechanisms underlying the downregulation of VEGF expression. An important finding was the observation that the loss of VEGF expression by podocytes and tubular cells was strongly correlated, both spatially and quantitatively, with macrophage infiltration. It is well documented that macrophages may have proinflammatory or reparative phenotypes and can express both proangiogenic and antiangiogenic factors (40). Macrophages may therefore be regarded as a “double-edged sword,” because they may be important in some angiogenic responses, such as those in wounds or tumors (40), but they also play an important role in renal damage and progressive scarring (41). Although some of the tubular loss of VEGF in our study could reflect tubular damage, this cannot account for the diffuse loss of VEGF from the outer medullary tubules. We therefore explored the possibility that key macrophage-derived cytokines, including IL-1β, IL-6, and TNF-α, may affect VEGF expression by tubular cells. Interestingly, IL-1β, IL-6, and TNF-α (albeit at a higher concentration) were all able to significantly inhibit the secretion of VEGF protein by mTAL cells in vitro and to suppress VEGF mRNA levels. The inhibitory effects of these cytokines on VEGF expression were also demonstrated under hypoxic conditions similar to those normally present in the outer medulla. These observations are the first to demonstrate that macrophage-associated cytokines can negatively modulate the constitutive production of VEGF by distal tubular epithelial cells. Inhibition of VEGF would be expected to negatively affect the ability of the microvasculature to respond to injury and thus represents another mechanism by which macrophages can contribute to progressive renal disease.

Proteinuria, which is well recognized as being an important risk factor for progression, was also correlated with capillary loss. Given the strong association of proteinuria with interstitial macrophage infiltration (42), it seems possible that proteinuria may indirectly contribute to microvascular loss, via stimulation of local macrophage accumulation.

In conclusion, the RK model is associated with a brief but unsustained angiogenic response, followed by progressive glomerular and peritubular capillary loss and renal scarring. The loss of capillaries is correlated with a loss of VEGF and an increase in TSP-1 levels in the kidney, conditions that favor endothelial cell loss and impaired angiogenesis. The observation that sites of macrophage accumulation were correlated with the loss of VEGF suggests that macrophages may play a role in the downregulation of VEGF. This proposal was supported by in vitro data demonstrating that macrophage-derived cytokines (IL-1β, IL-6, and TNF-α) significantly suppressed VEGF expression in mTAL cells. Therefore, progressive renal disease in this model is associated with impaired angiogenesis and capillary loss, suggesting a critical role for microvascular disease in either causing or contributing to the development of end-stage renal disease.

Support for this manuscript was provided by United States Public Health Service Grants DK43422, DK47659, and DK52121. Dr. Kang is the recipient of an International Society of Nephrology Fellowship Award, a COBE Research Grant, and a postdoctoral fellowship grant from the Korean Science and Engineering Foundation.

Raymond C. Harris served as guest editor and supervised the review and final disposition of this manuscript.

1. Bohle A, Mackensen-Haen S, Wehrmann M: Significance of post-glomerular capillaries in the pathogenesis of chronic renal failure. Kidney Blood Press Res 19:191 -195, 1996
2. Ohashi R, Kitamura H, Yamanaka N: Peritubular capillary injury during the progression of experimental glomerulonephritis in rats. J Am Soc Nephrol 11:47 -56, 2000
3. Thomas SE, Anderson S, Gordon KL, Oyama TT, Shankland SJ, Johnson RJ: Tubulointerstitial disease in aging: Evidence for peritubular capillary damage: A potential role for renal ischemia. J Am Soc Nephrol 9:231 -242, 1998
4. Epstein FH, Agmon Y, Brezis M: Physiology of renal hypoxia. Ann NY Acad Sci 718:72 -81, 1994
5. Falanga V, Kirsner RS: Low oxygen tension stimulates proliferation of fibroblasts seeded as single cells. J Cell Physiol154: 505-510,1993
6. Falanga V, Martin TA, Takagi R, Kirsner R, Helfman T, Pardes J, Ochoa MS: Low oxygen tension increases mRNA levels of alpha 1 (I) procollagen in human dermal fibroblasts. J Cell Physiol157: 408-412,1993
7. Orphanides C, Fine L, Norman JT: Hypoxia stimulates proximal tubular cell matrix production via a TGF-β-independent mechanism. Kidney Int 52:637 -647, 1997
8. Kitamura H, Shimizu A, Masuda Y, Ishizaki M, Sugisaki Y, Yamanaka N: Apoptosis in glomerular endothelial cells during the development of glomerulosclerosis in the remnant kidney model. Exp Nephrol 6:328 -336, 1998
9. Shimizu A, Kitamura H, Masuca Y, Ishizaki M, Sugisaki Y, Yamanaka N: Rare glomerular capillary regeneration and subsequent capillary regression with endothelial cell apoptosis in progressive glomerulonephritis. Am J Pathol 151:1231 -1239, 1997
10. Dimmeler S, Rippman V, Weiland U, Haendeler J, Zeiher A: Angiotensin II induces apoptosis of human endothelial cells. Circ Res 81: 970-976,1997
11. Ho FM, Liu SH, Liau CS, Huang PJ, Shiah SG, Lin-Shiau SY: Nitric oxide prevents apoptosis of human endothelial cells from high glucose exposure during early stage. J Cell Biochem75: 258-263,1999
12. Iruela-Arispe L, Gordon KL, Hugo C, Duijvestijn A, Claffey K, Reilly M, Couser WG, Alpers CE, Johnson RJ: Participation of the glomerular endothelial cell in capillary repair in glomerulonephritis. Am J Pathol 147:1715 -1727, 1995
13. Shimizu A, Matsuda Y, Kitamura J, Ishizaki M, Sugisaki Y, Yamanaka N: Recovery of damaged glomerular capillary network with endothelial cell apoptosis in experimental proliferative glomerulonephritis. Nephron 79:206 -214, 1998
14. Ibrahim HN, Hostetter T: The renin-aldosterone axis in two models of reduced renal mass in the rat. J Am Soc Nephrol9: 72-76,1998
15. Bunag RD, Buttefield J: Tail-cuff blood pressure measurement without external preheating in awake rats. Hypertension 4:898 -903, 1982
16. Kim YG, Suga S, Kang DH, Jefferson JA, Mazzali M, Gordon KL, Matsui K, Breiteneder-Geleff S, Shankland SJ, Hughes J, Kerjaschki D, Schreiner GF, Johnson RJ: Vascular endothelial growth factor stimulates vascular remodeling and tissue repair in a model of thrombotic microangiopathy. Kidney Int 58:2390 -2399, 2000
17. Duijvestijn AM, van Goor H, Klatter F, Majoor GD, van Bussel E, van Breda Vriesman PJ: Antibodies defining rat endothelial cells: RECA-1, a pan-endothelial cell-specific monoclonal antibody. Lab Invest 66:459 -466, 1992
18. Gerber HP, Hillan K, Ryan A, Kowalski J, Keller GA, Rangell L, Wright B, Radtke F, Aguet M, Ferrara N: VEGF is required for growth and survival in neonatal mice. Development126: 1149-1159,1999
19. Gibson UEM, Heid CA, Williams PM: A novel method for real time quantitative RT-PCR. Genome Res6: 995-1001,1996
20. Simon M, Grone H-J, Johren O, Kullmer J, Plate KH, Risau W, Fuchs E: Expression of vascular endothelial growth factor and its receptors in human renal ontogenesis and in adult kidney. Am J Physiol37: F24-F250,1995
21. Grone J-J, Simon M, Grone EF: Expression of vascular endothelial growth factor in renal vascular disease and renal allografts. J Pathol 177:259 -267, 1995
22. Floege JF, Burns MW, Alpers CE, Yoshimura A, Pritzl P, Gordon KL, Seifert RA, Bowen-Pope DF, Couser WG, Johnson RJ: Glomerular cell proliferation and PDGF expression precede glomerulosclerosis in the remnant kidney model. Kidney Int 41:297 -309, 1992
23. Kliem V, Johnson RJ, Alpers CE, Yoshimura A, Couser WG, Koch KM, Floege J: Mechanisms involved in the pathogenesis of tubulointerstitial fibrosis in 5/6-nephrectomized rats. Kidney Int49: 666-678,1996
24. Olivetti G, Anversa P, Rigamonti W, Vitali-Mazza L, Loud AV: Morphometry of the renal corpuscle during normal postnatal growth and compensatory hypertrophy: A light microscope study. J Cell Biol 75: 573-585,1977
25. Marcussen N, Nyengaard J, Christensen S: Compensatory growth of glomeruli is accomplished by an increased number of glomerular capillaries. Lab Invest 70:868 -874, 1994
26. Lee LK, Meyer TW, Pollock AS, Lovett DH: Endothelial cell injury initiates glomerular sclerosis in the rat remnant kidney. J Clin Invest 96:953 -964, 1995
27. Wu LL, Cox A, Roe CJ, Dziadek KM, Cooper ME, Gilbert RE: Transforming growth factor β and renal injury following subtotal nephrectomy in the rat: Role of the renin angiotensin system. Kidney Int 51:1553 -1567, 1997
28. Liu A, Dardik A, Ballermann BJ: Neutralizing TGF-betal antibody infusion in neonatal rat delays in vivo glomerular capillary formation. Kidney Int 56:1334 -1348, 1999
29. Ferrara N: Role of vascular endothelial growth factor in the regulation of angiogenesis. Kidney Int56: 794-814,1999
30. Spyridopoulos I, Brogi E, Kearney M, Sullivan AB, Cetrulo C, Isner JM, Losordo DW: Vascular endothelial growth factor inhibits endothelial cell apoptosis induced by tumor necrosis factor-alpha: Balance between growth and death signals. J Mol Cell Cardiol29: 1321-1330,1997
31. Gerber H-P, McMurtey A, Kowalski J, Yan M, Keyt BA, Dixit V, Ferrara N: Vascular endothelial growth factor regulates endothelial cell survival through the phosphatidylinositol 3′-kinase/Adkt signal transduction pathway. J Biol Chem273: 30336-30343,1998
32. Kang DH, Anderson S, Kim YG, Mazzali M, Suga S, Jefferson JA, Gordon KL, Oyama T, Hughes J, Hugo C, Kerjaschki D, Schreiner GF, Johnson RJ: Impaired angiogenesis in the aging kidney: Vascular endothelial growth factor and thrombospondin-1 in renal disease. Am J Kidney Dis37: 601-611,2001
33. Shulman K, Rosen S, Tognazzi K, Manseau EJ, Brown LF: Expression of vascular permeability factor (VPF/VEGF) is altered in many glomerular diseases. J Am Soc Nephrol 7:661 -666, 1996
34. Iruela-Arispe ML, Bornstein P, Sage H: Thrombospondin exerts an antiangiogenic effect on cord formation by endothelial cells in vitro.Proc Natl Acad Sci USA 88:5026 -5030, 1991
35. Guo N, Krutzsch HC, Inman JK, Roberts DD: Thrombospondin 1 and type 1 repeat peptides of thrombospondin 1 specifically induce apoptosis of endothelial cells. Cancer Res57: 1735-1742,1997
36. Abrass CK, Adcox MJ, Raugi GJ: Aging-associated changes in renal extracellular matrix. Am J Pathol146: 742-752,1995
37. Hugo C, Shankland S, Pichler R, Couser WG, Johnson RJ: Thrombospondin 1, a TGFβ activating protein, precedes and predicts the development of tubulointerstitial fibrosis in glomerular disease. Kidney Int 53:302 -311, 1998
38. McGregor B, Colon S, Mutin M, Chignier E, Zech P, McGregor J: Thrombospondin in human glomerulopathies: A marker of inflammation and early fibrosis. Am J Pathol 144:1281 -1287, 1994
39. Schultz-Cherry S, Ribeiro S, Gentry L, Murphy-Ullrich JE: Thrombospondin binds and activates the small and large forms of latent transforming growth factor-β in a chemically defined system. J Biol Chem 269:26775 -26782, 1994
40. Sunderkotter C, Steinbrink K, Goebeler M, Bhardwaj R, Sorg C: Macrophages and angiogenesis. J Leukocyte Biol55: 410-422,1994
41. Tesch GH, Schwarting A, Kinoshita K, Lan HY, Rollins BJ, Kelley VR: Monocyte chemoattractant protein-1 promotes macrophage-mediated tubular injury, but not glomerular injury, in nephrotoxic serum nephritis. J Clin Invest 103:73 -80, 1999
42. Eddy AA, Giachelli CM: Renal expression of genes that promote interstitial inflammation and fibrosis in rats with protein-overload proteinuria. Kidney Int 47:1546 -1557, 1995
Copyright © 2001 The Authors. Published by Wolters Kluwer Health, Inc. All rights reserved.