Acute Renal Failure in Endotoxemia Is Caused by TNF Acting Directly on TNF Receptor-1 in Kidney

Patrick N. Cunningham, Hristem M. Dyanov, Pierce Park, Jun Wang, Kenneth A. Newell and Richard J. Quigg
J Immunol June 1, 2002, 168 (11) 5817-5823; DOI: https://doi.org/10.4049/jimmunol.168.11.5817

Abstract

Bacterial endotoxin (LPS) is responsible for much of the widespread inflammatory response seen in sepsis, a condition often accompanied by acute renal failure (ARF). In this work we report that mice deficient in TNFR1 (TNFR1−/−) were resistant to LPS-induced renal failure. Compared with TNFR1+/+ controls, TNFR1−/− mice had less apoptosis in renal cells and fewer neutrophils infiltrating the kidney following LPS administration, supporting these as mediators of ARF. TNFR1+/+ kidneys transplanted into TNFR1−/− mice sustained severe ARF after LPS injection, which was not the case with TNFR1−/− kidneys transplanted into TNFR1+/+ mice. Therefore, TNF is a key mediator of LPS-induced ARF, acting through its receptor TNFR1 in the kidney.

Sepsis is a common and often devastating condition characterized by uncontrolled infection and multisystem organ failure (1). Bacterial endotoxin (LPS) is a cell wall component of Gram-negative bacteria that is responsible for much of the widespread inflammatory response seen in this condition. Administration of LPS to animals causes systemic hypotension and organ dysfunction similar to what occurs in clinical sepsis (2). Several mediators of inflammation have been assigned a key role in sepsis, among them TNF, which is released by macrophages and other cells into the circulation following LPS administration (3, 4). While experiments using TNF-blocking agents have shown protection against LPS-induced mortality in mice (5, 6, 7), clinical trials of similar agents in humans have failed to demonstrate any improvement in survival (8). TNF acts through two distinct receptors: TNFR1, which mediates most of the inflammatory actions of TNF, and TNFR2, which has been proposed to have an anti-inflammatory role (9). However, mice deficient in TNFR1 (TNFR1−/−) are not protected from mortality induced by administration of high doses of LPS (9, 10).

Sepsis is a leading cause of acute renal failure (ARF)3 (11, 12, 13), but the mechanisms behind this ARF remain unclear. Although cytokines such as TNF are key mediators of sepsis, little is known about their acute effects on the kidney in the setting of endotoxemia. We previously observed that, despite severe impairment of kidney function following LPS administration, the extent of renal pathologic injury is mild (14). To explain this dissociation between the severe functional impairment, yet the relatively mild pathologic injury seen in LPS-induced ARF, we hypothesized that apoptosis, which is difficult to detect by routine histologic studies, is responsible for renal cellular dysfunction and death. Of note, TNF has been shown to induce apoptosis in a variety of cell types, including renal tubular cells (15, 16). Therefore, we investigated the role and mechanisms of action of TNF in endotoxin-induced ARF.

Materials and Methods

Animals

Mice with a mutation in the Fas gene (B6.MRL-Faslpr; denoted Fas−/−) or with targeted deletion of TNFR1 (TNFR1−/−) or TNFR2 (TNFR2−/−) were obtained from The Jackson Laboratory (Bar Harbor, ME) and used for experiments at 8 wk of age. Because each of these genotypes is on a C57BL/6 background, age-matched C57BL/6 mice were used as controls. The animals were maintained and experiments were performed in accordance with the guidelines set by the University of Chicago Institutional Animal Care and Use Committee.

ARF model

Under isoflurane inhalational anesthesia, mice were given 0.15 mg Escherichia coli LPS (Sigma-Aldrich, St. Louis, MO) as an i.p. injection. Blood was obtained via retro-orbital bleeding immediately before injection and 24 h later, with sacrifice and harvest of kidney tissue and blood at 48 h after injection. Portions of kidney tissue were frozen in OCT compound (Fisher Scientific, Pittsburgh, PA) at −80°C, and also formalin-fixed and stained with periodic acid-Schiff base and H&E. Blood urea nitrogen (BUN) concentrations were determined with a Beckman CX5CE autoanalyzer (Beckman, Palo Alto, CA). At this relatively low dose of LPS, systemic blood pressure falls by <10 mm Hg in C57BL/6 mice (Ref. 17 and our unpublished observations). To evaluate the possibility that TNFR2−/− mice had an enhanced susceptibility to LPS-induced ARF, TNFR2−/− mice and their control group were given LPS at a lower dose of 0.075 mg, the rationale for which is described in greater detail below.

TUNEL staining

Mice were sacrificed at varying times after i.p. injection of 0.15 mg LPS, and kidney tissue was immediately frozen in OCT compound. TUNEL staining was performed with the TdT-FragEL DNA fragmentation detection kit (Oncogene, Boston, MA) according to the manufacturer’s instructions. In brief, 8-μm cryostat sections were washed in Tween 20/TBS and fixed in 4% formaldehyde for 30 min. This was followed by proteinase K treatment for 18 min and incubation in 3% H2O2/methanol for 7 min. Specimens were incubated with TdT at 37°C for 60 min, blocked in dilute BSA for 20 min, and incubated in streptavidin-HRP for 30 min, followed by detection with diaminobenzidine reagent for 15 min. Sections were counterstained with methyl green, and a blinded observer counted the number of positively stained nuclei per high-power field and recorded the average of 10 fields for each sample.

LM-PCR

Kidney tissue was harvested at various time points after injection of 0.15 mg LPS and immediately frozen at −80°C. Later, a small piece was thawed, from which DNA was purified (DNeasy DNA purification system; Qiagen, Valencia, CA). The extent of DNA laddering was amplified and detected by ligase-mediated (LM)-PCR (Clontech Laboratories, Palo Alto, CA) according to the manufacturer’s instructions, as follows. DNA isolated from each animal was incubated with the supplied primer targets and T4 DNA ligase for 18 h at 16°C. Twenty micrograms of this ligated DNA was then used as the substrate for PCR, using supplied primers and Advantage DNA polymerase (Clontech Laboratories) for 23 cycles of 94°C for 1 min and at 72°C for 3 min. The reaction product for each animal was run on a 1.3% agarose gel, and ethidium bromide-stained bands were detected with UV light illumination.

Kidney transplantation

A single kidney was transplanted from 10-wk-old TNFR1−/− to TNFR1+/+ mice, and vice versa. Each individual mouse was used only as a recipient or a donor. Mice were anesthetized with 65 mg/kg pentobarbital i.p. The left kidney of donor mice was perfused via the renal artery with 0.3 ml cold saline and resected with artery, vein, and ureter attached. The kidney was stored at 4°C until anastomosis. Next, the recipient underwent midline abdominal incision, followed by suprarenal clamping of aorta and inferior vena cava. The donor kidney was placed in the right flank, and its artery and vein were attached with 10-0 nylon suture via side-to-end anastomoses with the recipient aorta and inferior vena cava. The bladder was punctured with a 21-gauge needle, and the ureter was sewn in place. Native kidneys were then resected. Animals were allowed to recover for 10 days after surgery before LPS injection. Preliminary dosing experiments using syngeneic C57BL/6 kidney transplants showed that mouse kidney transplant recipients have a stable, mild elevation of BUN and creatinine but are somewhat more sensitive to LPS; thus, a dose of 0.1 mg was used in these studies.

Tissue processing

Histologic sections for each animal were assigned a semiquantitative score for tubular injury adopted from Nomura et al. (18). A blinded observer assigned a score ranging from 0 (no injury) to 3 (severe/widespread injury) for each of three variables: tubular dilatation/flattening, tubular casts, and tubular degeneration/vacuolization. For each animal, 10 cortical and 10 medullary high-power fields were examined at random. For each variable within each field a score of 0 was assigned when <5% of the tubules were affected, a score of 1 when 5–33% were affected, a score of 2 when 34–66% were affected, and a score of 3 when >66% were affected. The average score for each variable was calculated separately for the cortex and medulla, and all three variables from both cortex and medulla were summed to arrive at a total injury score for each animal ranging from 0 to 18.

For immunohistochemistry, 4-μm kidney cryostat sections were fixed with ether/ethanol, incubated with 0.3% H2O2 for 30 min, and blocked with dilute horse serum. Sections were stained for neutrophils by sequential incubation with rat anti-mouse neutrophil (mAb 7/4; Serotec, Raleigh, NC) at a 1/60 dilution for 30 min followed by HRP-conjugated rabbit anti-rat IgG (Sigma-Aldrich) at a 1/60 dilution for 30 min and diaminobenzidine reagent (Vector Laboratories, Burlingame, CA) for 10 min. A blinded observer counted the number of neutrophils per high-power field and recorded the average of 10 fields for each sample. Sections were stained for TNFR1 by incubation with hamster anti-mouse TNFR1 (BD PharMingen, San Diego, CA) at a 1/60 dilution for 30 min, followed by biotin-conjugated mouse anti-hamster IgG (BD PharMingen) at a 1/100 dilution for 30 min, streptavidin-HRP (Vector Laboratories) for 30 min, and diaminobenzidine reagent.

Measurement of TNF release

In a parallel set of experiments, blood was collected from wild-type C57BL/6 mice (n = 8) at baseline and again 2 h after LPS injection. Serum was analyzed for TNF levels using a commercially available TNF ELISA kit (R&D Systems, Minneapolis, MN) according to the manufacturer’s instructions.

mRNA transcript profiling in LPS-injected mice

For microarray studies, normal C57BL/6 mice were given 0.15 mg LPS, and 4 or 8 h later tissue was harvested and kidney RNA was isolated using TRIzol reagent (Life Technologies, Grand Island, NY). RNA from the kidney of a mouse injected with saline served as the control. RNA was processed and hybridized with murine genome U74A arrays (Affymetrix, Santa Clara, CA) according to the manufacturer’s instructions.

Real-time RT-PCR for inducible NO synthase (iNOS) was performed as follows. A portion of whole kidney from each TNFR1−/− and TNFR1+/+ mouse was frozen at −80°C, from which total RNA was purified using the RNeasy Mini RNA purification kit (Qiagen). To remove all traces of genomic DNA, samples were then treated with RNase-free RQ1 DNase (1 U per 4 μg RNA; Promega, Madison, WI) in 10 μl reaction buffer (final concentration, 40 mM Tris-HCl, 10 mM MgSO4, and 1 mM CaCl2 (pH 8)) at 37°C for 30 min. This was followed by addition of 1 μl 20 mM EGTA (pH 8) to stop the reaction and incubation at 65°C for 10 min to inactivate the DNase. cDNA was generated from RNA using random hexamers as primers with the SuperScript first-strand synthesis kit (Life Technologies), according to the manufacturer’s instructions. cDNA was prepared in an analogous fashion from kidneys of mice sacrificed 6 h after LPS injection and compared with sham-injected kidney cDNA for analysis of early TNF and TNFR1 mRNA (n = 6 per group).

Real-time PCR was performed using a Smart Cycler (Cepheid, Sunnyvale, CA). For iNOS and GAPDH measurements, probes were labeled at the 5′ end with the reporter dye molecule FAM (6-carboxyfluorescein; emission λmax = 518 nm) and at the 3′ end with the quencher dye molecule TAMRA (6-carboxytetramethyl-rhodamine; emission λmax = 582 nm). Each reaction was conducted in a total volume of 25 μl with 1× TaqMan Master Mix (PE Applied Biosystems, Foster City, CA), 3 μl sample or standard cDNA, primers at 200 nM each, and probe at 100 nM. PCR was conducted with a hot start at 95°C (5 min), followed by 45 cycles of 95°C for 15 s and 60°C for 30 s. For each sample, the number of cycles required to generate a given threshold signal (Ct) was recorded. Using a standard curve generated from serial dilutions of splenic cDNA, the ratio of iNOS expression relative to GAPDH expression was calculated for each experimental animal and normalized relative to an average of ratios from TNFR1+/+ kidney unexposed to LPS (n = 5). Measurements of TNF and TNFR1 mRNA expression relative to GAPDH were performed in an analogous fashion, but using the SybrGreen intercalating dye method with HotStar DNA polymerase (PE Applied Biosystems) instead of labeled probes. Reaction conditions were according to the manufacturer’s instructions. The reaction temperature profiles for TNF and TNFR1 were as described above, except that the annealing temperature was 58°C. Primers were synthesized by Integrated DNA Technologies (Coralville, IA) and probes by Synthegen (Houston, TX). The sequences of primers/probes were as follows: iNOS forward primer, 5′-CAGCTGGGCTGTACAAACCTT-3′; iNOS reverse primer, 5′-TGAATGTGATGTTTGCTTCGG-3′; iNOS probe, 5′-CGGGCAGCCTGTGAGACCTTTGA-3′; GAPDH forward primer, 5′-GGCAAATTCAACGGCACAGT-3′; GAPDH reverse primer, 5′-AGATGGTGATGGGCTTCCC-3′; GAPDH probe, 5′-AAGGCCGAGAATGGGAAGCTTGTCATC-3′; TNF forward primer, 5′-CCGATGGGTTGTACCTTGTC-3′; TNF reverse primer, 5′-GTGGGTGAGGAGCACGTAGT-3′; TNFR1 forward primer, 5′-GACCGGGAGAAGAGGGATAG-3′; and TNFR1 reverse primer, 5′-GTTCCTTTGTGGCACTTGGT-3′.

Statistics

Data were analyzed with Minitab software (State College, PA). Unless noted otherwise, data are given as the mean ± SEM. Groups were compared by two-sample t test. When comparing the severity of ARF between two groups, the slope of BUN vs time was determined for each individual animal by least squares regression, and the individual slopes in the two groups were compared with the Mann-Whitney rank-sum test. A value of p ≤ 0.05 was considered significant.

Results

Apoptosis occurs in the kidney after endotoxin administration

Several factors present in sepsis can serve as triggers of apoptosis, including hypoxia and reactive oxygen species (19, 20). Additionally, apoptosis can be induced in a variety of cell types, including renal tubular cells, in a receptor-mediated fashion, via Fas, TNFR1, and other TNFR family members (15, 16, 20, 21, 22, 23). To examine whether apoptosis occurs in the kidney after LPS administration, TUNEL staining was performed on kidney sections following administration of LPS to mice.

Normal mice had few TUNEL-stained nuclei, while LPS administration led to an increase in the number of apoptotic cells at time points ranging from 3 to 48 h later (0.8 ± 0.2 vs 2.3 ± 0.3 TUNEL-stained nuclei per high-power field at 8 h; p = 0.022; Fig. 1⇓). Most apoptotic cells appeared to be tubular epithelial cells. To further quantify the extent of apoptosis, LM-PCR was performed to amplify fragmented DNA from kidney specimens. This demonstrated that the limited extent of apoptosis present in normal kidney was increased in a time-dependent fashion following LPS administration (Fig. 2⇓).

FIGURE 1.
FIGURE 1.

Apoptosis occurs in the kidney after LPS administration. A, TUNEL-stained apoptotic nuclei are rare in normal mouse kidney. B, At time points ranging from 3 to 48 h after LPS administration, darkly staining apoptotic nuclei were visible at an increased frequency (arrows). Apoptotic nuclei were predominantly observed in the tubules. Kidney harvested 8 h after LPS injection is shown. Methyl green counterstain; magnification, ×400.

FIGURE 2.
FIGURE 2.

Time course of apoptosis in the kidney after LPS administration. DNA was isolated from C57BL/6 mice at various time points after injection of 0.15 mg LPS and subjected to LM-PCR. The small amount of apoptosis occurring in kidney at baseline (BL) is markedly increased following LPS administration, peaks at 8–12 h, and is still apparent 48 h after injection.

Mice deficient in TNFR1 are resistant to endotoxin-induced ARF

In addition to the release of TNF into the circulation that occurs after LPS administration, Fas and Fas ligand have been shown to be up-regulated on the surface of renal tubular cells following LPS administration in vivo (24). Thus, we hypothesized that one or both of these pathways could account for renal cell apoptosis and ARF following LPS administration. Fas−/− mice developed ARF to the same extent as Fas+/+ controls following LPS injection (Fig. 3⇓A). In contrast, TNFR1−/− mice were resistant to LPS-induced ARF (Fig. 3⇓B; n = 8 per group; p = 0.005 compared with TNFR1+/+ mice). In addition to protection from renal functional impairment, TNFR1−/− mice had less renal tubular injury (tubular injury scores of 3.9 ± 0.6 vs 7.1 ± 1.1 in TNFR1−/− and TNFR1+/+ mice, respectively; p = 0.027; Fig. 4⇓). Therefore, TNF, acting through TNFR1, is a key mediator of LPS-induced ARF. Renal apoptosis, as assessed by LM-PCR, was greater in the TNFR1+/+ controls than in the TNFR1−/− group. Furthermore, the amount of apoptosis in individual animals correlated with the extent of renal functional impairment (Fig. 5⇓), suggesting that apoptosis is intimately involved in LPS-induced ARF.

FIGURE 3.
FIGURE 3.

TNFR1−/− mice are resistant to LPS-induced ARF. Renal function, as measured by BUN, is shown at 24 and 48 h following LPS injection. A, Both Fas−/− and Fas+/+ mice developed ARF following injection of 0.15 mg LPS (n = 10 in each group). B, TNFR1−/− mice did not develop ARF following LPS administration, in marked contrast to TNFR1+/+ control mice (n = 8 in each group; p = 0.005). C, Both TNFR2−/− and TNFR2+/+ mice developed ARF after injection of 0.075 mg LPS (n = 8 in each group).

FIGURE 4.
FIGURE 4.

Renal histology in LPS-induced ARF. A, TNFR1+/+ mice developed a modest extent of tubular injury, including tubular dilatation and flattening (*) and vacuolization of renal tubular cells (arrows). B, In contrast, there was significantly less tubular damage in TNFR1−/− mice. The sections shown were harvested 48 h after injection of 0.15 mg LPS and stained with H&E. Magnification, ×400.

FIGURE 5.
FIGURE 5.

Quantification of apoptosis occurring in kidney after LPS injection. In both groups there was a correlation between the functional extent of renal failure, as indicated by 48-h BUN values, and the extent of apoptosis. A, LM-PCR shows a strong amount of laddering in TNFR1+/+ mice. Each lane contains kidney tissue harvested from an individual animal 48 h after injection of LPS. BL, Sample of normal C57BL/6 kidney DNA at baseline, not exposed to LPS. B, In contrast, the amount of kidney apoptosis after LPS was markedly less in TNFR1−/− mice. Calf thymus DNA was used as a positive control (+).

The role of TNFR2 in apoptosis and inflammation is less clear. While apoptotic signaling has been clearly demonstrated to occur through TNFR2 (25), an antiapoptotic role has also been ascribed to TNFR2 (26, 27). Also, soluble TNFR2 may serve to down-regulate TNF action (9). Thus, we thought it conceivable that TNFR2−/− mice would be more sensitive than TNFR2+/+ mice to LPS-induced ARF. However, using a lower dose of 0.075 mg LPS, TNFR2−/− mice developed ARF to a similar extent as TNFR2+/+ controls, as assessed by BUN levels (Fig. 3⇑C) and tubular injury scores (5.6 ± 0.3 vs 5.3 ± 0.3 in TNFR2−/− and TNFR2+/+ mice, respectively; p = 0.42). By LM-PCR, TNFR2−/− and TNFR2+/+ mice had a similar amount of apoptosis in the kidney after LPS administration (data not shown). Therefore, TNFR2 does not appear to have a role in mediating the ARF or renal apoptosis induced by LPS.

TNFR1 mediates LPS-induced neutrophil infiltration in the kidney

In addition to apoptosis, there are other mechanisms by which TNFR1 could mediate LPS-induced renal injury. Although relatively mild, the pathologic changes seen in TNFR1+/+ mice after LPS administration cannot be explained by the noninflammatory process of apoptosis. TNF has been shown to stimulate the oxidative burst of neutrophils (28, 29, 30) and to up-regulate the expression of chemokines (31) and adhesion molecules (32, 33, 34) that mediate neutrophil recruitment to tissues following an inflammatory stimulus. Neutrophils have been implicated in the pathogenesis of LPS-induced injury of the liver and lung (35, 36, 37) and in ischemia-reperfusion injury of the kidney (38, 39). We confirmed that neutrophils infiltrated the kidney following LPS administration, but to a significantly lesser extent in TNFR1−/− mice compared with TNFR1+/+ controls (5.4 ± 0.8 vs 9.0 ± 1.3 neutrophils per high-power field in TNFR1−/− and TNFR1+/+ mice, respectively; p = 0.039; Fig. 6⇓). Additionally, the extent of renal neutrophil accumulation correlated with the degree of renal functional impairment in individual animals, as measured by BUN levels (r = 0.89; p < 0.001), further supporting a role for TNF-mediated neutrophil recruitment in the pathogenesis of LPS-induced ARF.

FIGURE 6.
FIGURE 6.

Neutrophil accumulation in the kidney 48 h after LPS injection. A, Few neutrophils are present in TNFR1+/+ kidney at baseline. B, Following injection of LPS, marked neutrophil accumulation was found in the kidneys of TNFR1+/+ mice, primarily in the renal cortex, which is shown. C, Although TNFR1−/− mice had an increase in neutrophil accumulation in the kidney after injection of LPS, it was significantly less than that in TNFR1+/+ mice. Magnification, ×200.

LPS causes up-regulation of iNOS expression independently of TNFR1

NO is a key mediator of a wide range of inflammatory as well as anti-inflammatory processes and has been shown to have direct cytotoxic effects on renal tubular cells (40) as well as complex actions on the renal vasculature (41). Inducible NOS has been shown to be strongly up-regulated in a wide variety of tissues, including the kidney, after administration of LPS (42). By real-time RT-PCR, expression of iNOS mRNA in the kidney was markedly up-regulated in both TNFR1−/− and TNFR1+/+ mice given LPS compared with the baseline value, but with no statistical differences between the two groups (4.7 ± 1.3- and 8.7 ± 3.0-fold increases in TNFR1−/− and TNFR1+/+ mice, respectively; p = 0.25; Fig. 7⇓). Furthermore, iNOS expression did not correlate with BUN levels (r = 0.326; p = 0.24). Thus, despite induction of iNOS mRNA as in wild-type controls, TNFR1−/− mice are protected from LPS-induced ARF.

FIGURE 7.
FIGURE 7.

Quantification of iNOS mRNA in kidney after LPS injection by real-time RT-PCR. Following injection of LPS, iNOS mRNA was up-regulated in the kidney in both groups compared to baseline, to a statistically similar extent. Each data point represents an individual animal.

TNF mediates ARF by acting on its receptor in kidney

As has been documented by others, the injection of LPS led to profound release of TNF into the circulation within 2 h (0.02 ± 0.02 ng/ml at baseline vs 2.70 ± 0.65 ng/ml 2 h after LPS; p < 0.001). The release of TNF soon after LPS administration is responsible for the later appearance of other circulating cytokines (43, 44). Also, injection of TNF in vivo duplicates the severe hypotension seen in septic shock (45). Thus, TNF could mediate LPS-induced ARF indirectly, by inducing systemic hypotension and/or by leading to the release of other inflammatory mediators into the circulation that could act on the kidney. Alternatively, LPS-induced ARF could be mediated by TNF acting directly on renal TNFR1. In support of the latter possibility, kidney sections from normal C57BL/6 mouse kidney stained positively for TNFR1, predominantly in cortical tubules (data not shown). In addition, using a microarray approach, we found TNFR1 mRNA to be up-regulated in the kidney by factors of 2.3- and 5.9-fold at 4 and 8 h after LPS administration, respectively. These preliminary data were confirmed in subsequent experiments using real-time RT-PCR, showing that TNFR1 mRNA was up-regulated 3.0 ± 0.2-fold in kidney at 6 h after LPS administration in comparison with sham-injected controls (n = 6 per group; p < 0.001).

To address the question of the level of action of TNF, we transplanted kidneys between TNFR1+/+ and TNFR1−/− mice and examined their responses to LPS. TNFR1+/+ kidneys in TNFR1−/− mice developed severe ARF after LPS administration, which was not the case for TNFR1−/− kidneys in TNFR1+/+ recipients (n = 4 per group; p = 0.029; Fig. 8⇓). As with native kidneys in TNFR1−/− mice, pathologic specimens of transplanted TNFR1−/− kidneys had less injury than transplanted TNFR1+/+ kidneys (tubular injury scores of 3.6 ± 0.6 vs 5.8 ± 0.4; p = 0.062). These data prove that TNF causes LPS-induced ARF through its action on TNFR1 in the kidney. In addition to its systemic release, TNF was also strongly up-regulated in the kidney (14.6 ± 3.5-fold increase 6 h after LPS injection; n = 6; p < 0.001) as assessed by real time RT-PCR, suggesting that not only systemic but also locally produced TNF was acting through TNFR1 to cause LPS-induced ARF, as has been previously suggested (46). TNFR1−/− transplanted kidneys had significantly less neutrophil accumulation (3.9 ± 1.4 vs 12.0 ± 0.7 PMNs/high-power field in TNFR1−/− and TNFR+/+ kidneys, respectively; p = 0.015) as well as a lesser extent of apoptosis as quantified by LM-PCR (data not shown), further strengthening a role for both apoptosis and neutrophil-mediated injury in the pathogenesis of LPS-induced ARF.

FIGURE 8.
FIGURE 8.

TNF acts directly on the kidney to cause ARF after LPS injection. BUN values are shown as indicators of renal function at various times following LPS injection. TNFR1+/+ kidneys in TNFR1−/− mice (TNFR1+/+→TNFR1−/−; n = 4) developed marked ARF after injection of LPS. In contrast, TNFR1−/− kidneys in TNFR1+/+ mice (TNFR1−/−→TNFR1+/+, n = 4) were protected from ARF after LPS administration. p = 0.029 comparing the two groups.

Discussion

The pathogenesis of sepsis and the associated end-organ failure is highly complex, involving multiple inflammatory mediators (2). Although an important role for TNF in sepsis has been established, the role of TNF in LPS-induced ARF has only recently been examined. Knotek et al. (17) found that the administration of a TNF-binding protein protected mice against LPS-induced ARF, but they did not find evidence of renal apoptosis or neutrophil infiltration described herein. In this work we show that LPS-induced ARF can be attributed to TNF acting directly on its receptor, TNFR1, in kidney.

The multisystem organ failure that occurs in sepsis is often ascribed to widespread hypotension and impairment of oxygen delivery. This explanation is especially intuitive in the kidney, where hemodynamic forces drive glomerular filtration, and a high metabolic activity exists in the transporting tubular epithelium (47, 48). TNF is well known to lead to the release of multiple other cytokines into the circulation (43, 44). Thus, we posed the question of whether TNF mediates LPS-induced ARF indirectly, by leading to hypotension and/or the release of other inflammatory mediators into the circulation. However, our data showing severe LPS-induced ARF in TNFR1+/+ kidneys transplanted into TNFR1−/− recipients, with relative protection of TNFR1−/− kidneys in TNFR1+/+ recipients, prove that TNF mediates LPS-induced ARF directly in the kidney. In addition, the fact that TNFR1−/− mice are fully sensitive to LPS-induced mortality (9, 10) yet resistant to LPS-induced ARF further supports a renal-specific role for TNF.

A leading experimental model of ARF is that of ischemia-reperfusion, in which the renal artery is completely occluded for a period of time and then released. However, this is not an accurate simulation of what occurs in most cases of human ARF. Instead, sepsis with attendant multisystem organ failure is a leading cause of human ARF in most series (11, 12). Most pathologic specimens from human ARF show the same relatively mild degree of histologic damage that we note in our model (49, 50). Therefore, the conclusions we present here are especially relevant to ARF in humans.

Given the complexity of sepsis, the pathogenesis of LPS-induced ARF is probably multifactorial. Our studies point to two possible mechanisms by which TNF causes ARF: renal cell apoptosis and neutrophil infiltration. Apoptosis is increasingly recognized to play a key role in ischemia-reperfusion injury in the kidney and other organs (21, 51), and inhibition of apoptosis with caspase inhibitors has been shown to protect against ischemia-reperfusion renal injury (52). Overall, given our findings and the fact that TNF acting on TNFR1 is a well-described trigger of apoptosis (15, 16), it seems likely that in the setting of LPS-induced TNF release (3, 4), renal TNFR1 is activated and results in apoptosis. However, other known triggers of apoptosis, such as hypoxia or reactive oxygen intermediates, the latter perhaps derived from infiltrating neutrophils, which are present in the setting of sepsis could be relevant as well. Further delineation of the precise apoptotic pathways involved in LPS-induced ARF is the subject of ongoing work. A role for neutrophils in LPS- and ischemia-reperfusion-induced ARF has also been suggested by others (29, 38, 39). Our findings that TNFR1−/− mice given LPS were protected from ARF and had fewer neutrophils infiltrating the kidney are consistent with these previous observations. Possible mechanisms for the recruitment of neutrophils to the kidney include TNF acting through TNFR1 to up-regulate adhesion molecules such as E- and P-selectin and ICAM-1 (32, 33, 34) or chemotactic molecules such as IFN-inducible protein-10 and macrophage inflammatory protein-2 (31). The fact that apoptosis and neutrophil accumulation each correlated with the degree of LPS-induced renal failure in individual animals supports a causative role, but additional experiments must be conducted to determine the extent to which these mechanisms individually contribute to LPS-induced ARF and the extent to which these processes are interrelated.

NO has received great attention as an important mediator in ARF. Its role is complex, with both noxious and protective effects (53). Specific inhibition of iNOS, which is markedly up-regulated in the kidney after LPS administration, but not broad-spectrum NOS inhibition, has been shown to be protective against LPS-induced ARF, implying that it is the inducible isoform of NO synthase that mediates injury in this disease model (41). Because TNF has been shown to up-regulate iNOS in a variety of cell types, including rat proximal tubule cells (54), we hypothesized that the TNFR1−/− mice were resistant to LPS-induced ARF in part because of a lesser induction of iNOS. However, our finding that iNOS was up-regulated to a similar extent in both TNFR1−/− and TNFR1+/+ kidneys shows that signaling through TNFR1 is not required for iNOS transcription and that the resistance to LPS-induced ARF which we observed in TNFR1−/− mice must be due to other mechanisms. This finding that iNOS has little role in LPS-induced ARF in mice is in agreement with the observations of Knotek et al. (17) that iNOS−/− mice are still susceptible to LPS-induced ARF.

Clinical sepsis is a complex phenomenon, due to a variety of bacterial products released intermittently or continuously, often in the setting of considerable comorbidities, and is therefore imperfectly simulated by animal models. Nevertheless, the key role of bacterial endotoxin in severe sepsis is undisputed. Although human trials of TNF-blocking agents in septic shock have been disappointing because they do not appear to affect mortality (8), prevention of renal failure in septic patients receiving anti-TNF Ab has been observed (55). Taken together, these data in humans and mice are consistent and illustrate the predominant role of TNF in the ARF of sepsis. Further definition of the effects of TNF on the kidney will lead to a better understanding of the pathophysiology of ARF occurring during the course of sepsis.

Footnotes

  • ↵1 This work was supported by National Institutes of Health Grants R01DK41873, R01DK55357, and R01AI43579 and grants from the National Center and Greater Chicago Chapter of the Arthritis Foundation. P.N.C. and P.P. were supported by National Institutes of Health Training Grant T32DK07510.

  • ↵2 Address correspondence and reprint requests to Dr. Patrick N. Cunningham, Section of Nephrology, MC5100, University of Chicago, 5841 South Maryland Avenue, Chicago, IL 60637. E-mail address: pcunning{at}medicine.bsd.uchicago.edu

  • ↵3 Abbreviations used in this paper: ARF, acute renal failure; BUN, blood urea nitrogen; iNOS, inducible NO synthase; LM, ligase-mediated.

  • Received November 9, 2001.
  • Accepted March 6, 2002.

References

  1. ↵
    Young, L.. 1995. Sepsis syndrome. G. Mandel, and J. Bennet, and R. Dolan, eds. Principles and Practice of Infectious Diseases 806 Churchill Livingston, New York.
  2. ↵
    Bone, R. C.. 1991. The pathogenesis of sepsis. Ann. Intern. Med. 115: 457
    OpenUrlCrossRefPubMed
  3. ↵
    Beutler, B. A., I. W. Milsark, A. Cerami. 1985. Cachectin/TNF: production, distribution, and metabolic fate in vivo. J. Immunol. 135: 3972
    OpenUrlAbstract
  4. ↵
    van Deventer, S. J., H. R. Buller, J. W. ten Cate, L. A. Aarden, C. E. Hack, A. Sturk. 1990. Experimental endotoxemia in humans: analysis of cytokine release and coagulation, fibrinolytic, and complement pathways. Blood 76: 2520
    OpenUrlAbstract/FREE Full Text
  5. ↵
    Ashkenazi, A., S. A. Marsters, D. J. Capon, S. M. Chamow, I. S. Figari, D. Pennica, D. V. Goeddel, M. A. Palladino, D. H. Smith. 1991. Protection against endotoxic shock by a tumor necrosis factor receptor immunoadhesin. Proc. Natl. Acad. Sci. USA 88: 10535
    OpenUrlAbstract/FREE Full Text
  6. ↵
    Mohler, K. M., D. S. Torrance, C. A. Smith, R. G. Goodwin, K. E. Stemler, V. P. Fung, H. Madani, M. B. Widmer. 1993. Soluble tumor necrosis factor (TNF) receptors are effective therapeutic agents in lethal endotoxemia and function simultaneously as both TNF carriers and TNF antagonists. J. Immunol. 151: 1548
    OpenUrlAbstract
  7. ↵
    Tracey, K. J., Y. M. Fong, D. G. Hesse, K. R. Manogue, A. T. Lee, G. C. Kno, S. F. Lowry, A. Cerami. 1987. Anti-cachectin/TNF monoclonal Abs prevent septic shock during lethal bacteremia. Nature 330: 662
    OpenUrlCrossRefPubMed
  8. ↵
    Porter, S. B.. 2000. Current status of clinical trials with anti-TNF. Chest 112: 339
    OpenUrl
  9. ↵
    Peschon, J. J., D. S. Torrance, K. L. Stocking, M. B. Glaccum, C. Otten, C. R. Willis, K. Charrier, P. J. Morrissey, C. B. Ware, K. M. Mohler. 2000. TNF receptor-deficient mice reveal divergent roles for p55 and p75 in several models of inflammation. J. Immunol. 160: 943
    OpenUrlAbstract/FREE Full Text
  10. ↵
    Rothe, J., W. Lesslauer, H. Lotscher, Y. Lang, P. Koebel, F. Kontgen, A. Althage, R. Zinkernagel, M. Steinmetz, H. Bluethmann. 1993. Mice lacking the tumor necrosis factor receptor 1 are resistant to TNF-mediated toxicity but highly susceptible to infection by Listeria monocytogenes. Nature 364: 798
    OpenUrlCrossRefPubMed
  11. ↵
    Groenveld, A. B., D. D. Tran, J. van der Meulen, J. J. Nauta, L. G. Thijs. 1991. Acute renal failure in the medical intensive care unit: predisposing, complicating factors and outcome. Nephron. 59: 602
    OpenUrlCrossRefPubMed
  12. ↵
    Neveu, H., D. Kleinknecht, F. Brivet, P. Loirat, P. Landais. 1996. Prognostic factors in acute renal failure due to sepsis: results of a prospective multicentre study: the French Study Group on Acute Renal Failure. Nephrol. Dial. Transplant. 11: 293
    OpenUrlAbstract/FREE Full Text
  13. ↵
    Wardle, N.. 1982. Acute renal failure in the 1980s: the importance of septic shock and of endotoxaemia. Nephron 30: 193
    OpenUrlCrossRefPubMed
  14. ↵
    Cunningham, P. N., V. M. Holers, J. J. Alexander, J. M. Guthridge, M. C. Carroll, R. J. Quigg. 2000. Complement is activated in kidney by endotoxin but does not cause the ensuing acute renal failure. Kidney Int. 58: 1580
    OpenUrlCrossRefPubMed
  15. ↵
    Glynne, P. A., T. J. Evans. 2000. Inflammatory cytokines induce apoptotic and necrotic cell shedding from human proximal tubular epithelial cell monolayers. Kidney Int. 55: 2573
    OpenUrl
  16. ↵
    Peralta-Solter, A. P., J. M. Mullin, K. A. Knudsen, C. W. Marano. 1996. Tissue remodeling during tumor necrosis factor-induced apoptosis in LLC-PK1 renal epithelial cells. Am. J. Physiol. 270: F869
    OpenUrlAbstract/FREE Full Text
  17. ↵
    Knotek, M., B. Rogachev, W. Wang, T. Ecder, V. Melnikov, P. E. Gengaro, M. Esson, C. L. Edelstein, C. A. Dinarello, R. W. Schrier. 2001. Endotoxemic renal failure in mice: role of tumor necrosis factor independent of inducible nitric oxide synthase. Kidney Int. 59: 2243
    OpenUrlCrossRefPubMed
  18. ↵
    Nomura, A., K. Nishikawa, Y. Yuzawa, H. Okada, N. Okada, B. P. Morgan, S. J. Piddlesden, M. Nadai, T. Hasegawa, S. Matsuo. 1995. Tubulointerstitial injury induced in rats by a monoclonal Ab which inhibits function of a membrane inhibitor of complement. J. Clin. Invest. 96: 2348
    OpenUrlCrossRefPubMed
  19. ↵
    Lieberthal, W., S. A. Menza, J. S. Levine. 1998. Graded ATP depletion can cause necrosis or apoptosis of cultured mouse proximal tubular cells. Am. J. Physiol. 274: F315
    OpenUrlAbstract/FREE Full Text
  20. ↵
    Saikumar, P., Z. Dong, V. Mikhailov, M. Denton, J. M. Weinberg, M. A. Venkatachalam. 1999. Apoptosis: definition, mechanisms, and relevance to disease. Am. J. Med. 107: 489
    OpenUrlCrossRefPubMed
  21. ↵
    Lieberthal, W., J. S. Koh, J. S. Levine. 1998. Necrosis and apoptosis in acute renal failure. Semin. Nephrol. 18: 505
    OpenUrlPubMed
  22. ↵
    Ortiz-Arduan, O., T. M. Danoff, R. Kalluri, S. Gonzalez-Cuadrado, S. L. Karp, K. Elkton, J. Egido, E. G. Neilson. 1996. Regulation of Fas and Fas ligand expression in cultured murine renal cells and in the kidney during endotoxemia. Am. J. Physiol. 271: F1193
    OpenUrlAbstract/FREE Full Text
  23. ↵
    Ueda, N., G. P. Kaushal, S. V. Shah. 2000. Apoptotic mechanisms in acute renal failure. Am. J. Med. 108: 403
    OpenUrlCrossRefPubMed
  24. ↵
    Koide, N., K. Narita, Y. Kato, T. Sugiyama, D. Chakravortty, T. Yoshida, T. Yokochi. 1999. Expression of Fas and Fas ligand on mouse renal tubular epithelial cells in the generalized Shwartzman reaction and its relationship to apoptosis. Infect. Immun. 78: 4112
    OpenUrl
  25. ↵
    Beyaert, R., B. Vanhaesebroeck, W. Declercq, J. Van Lint, P. Vandenabele, P. Agostinis, J. R. Vandenheede, W. Fiers. 1995. Casein kinase-1 phosphorylates the p75 tumor necrosis factor receptor and negatively regulates tumor necrosis factor signaling for apoptosis. J. Biol. Chem. 270: 29293
    OpenUrlAbstract/FREE Full Text
  26. ↵
    Funk, J. O., H. Walczak, C. Voigtlander, S. Berchtold, T. Baumeister, P. Rauch, S. Rossner, A. Steinkasserer, G. Schuler, M. B. Lutz. 2000. Resistance to apoptosis and continuous proliferation of dendritic cells deficient for TNF receptor-1. J. Immunol. 165: 4792
    OpenUrlAbstract/FREE Full Text
  27. ↵
    Koch, F., C. Heufler, E. Kampgen, D. Schneeweiss, G. Bock, G. Schuler. 1990. Tumor necrosis factor α maintains the viability of murine epidermal Langerhans cells in culture, but in contrast to granulocyte/macrophage colony-stimulating factor, without inducing their functional maturation. J. Exp. Med. 171: 159
    OpenUrlAbstract/FREE Full Text
  28. ↵
    Dusi, S., B. Della, M. V, K. A. Donini, K. A. Nadalini, F. Rossi. 1996. Mechanisms of stimulation of the respiratory burst by TNF in nonadherent neutrophils: its independence of lipidic transmembrane signaling and dependence on protein tyrosine phosphorylation and cytoskeleton. J. Immunol. 157: 4615
    OpenUrlAbstract
  29. ↵
    Linas, S. L., D. Whittenburg, J. E. Repine. 1991. Role of neutrophil derived oxidants and elastase in lipopolysaccharide-mediated renal injury. Kidney Int. 39: 618
    OpenUrlCrossRefPubMed
  30. ↵
    Niwa, M., O. Kozawa, H. Matsuno, Y. Kanamori, A. Hara, T. Uematsu. 2000. Tumor necrosis factor-α-mediated signal transduction in human neutrophils: involvement of sphingomyelin metabolites in the priming effect of TNF-α on the fMLP-stimulated superoxide production. Life Sci. 66: 245
    OpenUrlCrossRefPubMed
  31. ↵
    Maier, S., K. Emmanuilidis, M. Entleutner, N. Zantl, M. Werner, K. Pfeffer, C. D. Heidecke. 2000. Massive chemokine transcription in acute renal failure due to polymicrobial sepsis. Shock 14: 187
    OpenUrlCrossRefPubMed
  32. ↵
    Koide, N., K. Abe, K. Narita, Y. Kato, T. Sugiyama, T. Yoshida, T. Yokochi. 1997. Expression of intercellular adhesion molecule-1 (ICAM-1) on vascular endothelial cells and renal tubular cells in the generalized Shwartzman reaction as an experimental disseminated intravascular coagulation model. FEMS Immunol. Med. Microbiol. 18: 67
    OpenUrlAbstract/FREE Full Text
  33. ↵
    Neumann, B., T. Machleidt, A. Lifka, K. Pfeffer, D. Vestweber, T. W. Mak, B. Holzmann, M. Kronke. 1996. Crucial role of 55-kilodalton TNF receptor in TNF-induced adhesion molecule expression and leukocyte organ infiltration. J. Immunol. 156: 1587
    OpenUrlAbstract
  34. ↵
    Weller, A., S. Isenmann, D. Vestweber. 1992. Cloning of the mouse endothelial selectins: expression of both E- and P-selectin is inducible by tumor necrosis factor α. J. Biol. Chem. 267: 15176
    OpenUrlAbstract/FREE Full Text
  35. ↵
    Abraham, E., A. Carmody, R. Shenkar, J. Arcaroli. 2000. Neutrophils as early immunologic effectors in hemorrhage- or endotoxemia-induced acute lung injury. Am. J. Physiol. 279: L1137
    OpenUrlAbstract/FREE Full Text
  36. ↵
    Essani, N. A., M. A. Fisher, C. A. Simmons, J. L. Hoover, A. Farhood, H. Jaeschke. 1998. Increased P-selectin gene expression in the liver vasculature and its role in the pathophysiology of neutrophil-induced liver injury in murine endotoxin shock. J. Leukocyte Biol. 63: 288
    OpenUrlAbstract
  37. ↵
    Kamochi, M., F. Komochi, Y. B. Kim, S. Sawh, J. M. Sanders, I. Sarembock, S. Green, J. S. Young, K. Ley, S. M. Fu, et al 1999. P-selectin and ICAM-1 mediate endotoxin-induced neutrophil recruitment and injury to the lung and liver. Am. J. Physiol. 277: L310
    OpenUrlAbstract/FREE Full Text
  38. ↵
    Kelly, K. J., W. W. Williams, Jr, R. B. Colvin, S. M. Meehan, T. A. Springer, J.-C. Gutierrez-Ramos, J. V. Bonventre. 2000. Intercellular adhesion molecule-1-deficient mice are protected against ischemic renal injury. J. Clin. Invest. 97: 1056
    OpenUrlCrossRefPubMed
  39. ↵
    Takada, M., K. C. Nadeau, G. D. Shaw, K. A. Marquette, N. L. Tilney. 1997. The cytokine-adhesion molecule cascade in ischemia/reperfusion injury of the rat kidney: inhibition by a soluble P-selectin ligand. J. Clin. Invest. 99: 2682
    OpenUrlCrossRefPubMed
  40. ↵
    Kabore, A. F., M. Denis, M. G. Bergeron. 1997. Association of nitric oxide production by kidney proximal tubular cells in response to lipopolysaccharide and cytokines with cellular damage. Antimicrob. Agents Chemother. 41: 557
    OpenUrlAbstract/FREE Full Text
  41. ↵
    Schwartz, D., M. Mendonca, I. Schwartz, Y. Xia, J. Satriano, C. B. Wilson, R. C. Blantz. 1997. Inhibition of constitutive nitric oxide synthase (NOS) by nitric oxide generated by inducible NOS after lipopolysaccharide administration provokes renal dysfunction in rats. J. Clin. Invest. 100: 439
    OpenUrlCrossRefPubMed
  42. ↵
    Morrissey, J. J., R. McCracken, H. Kaneto, M. Vehaskari, D. Montani, S. Klahr. 1994. Location of an inducible nitric oxide synthase mRNA in the normal kidney. Kidney Int. 45: 998
    OpenUrlCrossRefPubMed
  43. ↵
    Blackwell, T. S., J. W. Christman. 1996. Sepsis and cytokines: current status. Br. J. Anaesth. 77: 110
    OpenUrlAbstract/FREE Full Text
  44. ↵
    Tracey, K., A. Cerami. 1993. Tumor necrosis factor: an updated review of its biology. Crit. Care Med. 21: S415
    OpenUrlPubMed
  45. ↵
    Kilbourn, R. G., S. S. Gross, A. Jubran, J. Adams, O. W. Griffith, R. Levi, R. F. Lodato. 1990. NG-methyl-l-arginine inhibits tumor necrosis factor-induced hypotension: implications for the involvement of nitric oxide. Proc. Natl. Acad. Sci. USA 87: 3629
    OpenUrlAbstract/FREE Full Text
  46. ↵
    Donnahoo, K. K., X. Meng, L. Ao, A. Ayala, B. D. Shames, M. P. Cain, A. H. Harken, D. R. Meldrum. 2001. Differential cellular immunolocalization of renal tumour necrosis factor-α production during ischaemia versus endotoxaemia. Immunology 102: 53
    OpenUrlCrossRefPubMed
  47. ↵
    Bock, H. A.. 1998. Pathophysiology of acute renal failure in septic shock: from prerenal to renal failure. Kidney Int. 64: S15
    OpenUrl
  48. ↵
    Groeneveld, A. B.. 1994. Pathogenesis of acute renal failure during sepsis. Nephrol. Dial. Transplant. 9: (Suppl. 4):47
    OpenUrlAbstract/FREE Full Text
  49. ↵
    Racusen, L. C., B. A. Fivush, Y. L. Li, I. Slatnick, K. Solez. 1991. Dissociation of tubular cell detachment and tubular cell death in clinical and experimental “acutetubular necrosis.”. Lab. Invest. 64: 546
    OpenUrlPubMed
  50. ↵
    Solez, K., L. Morel-Maroger, J. D. Sraer. 1979. The morphology of “acute tubular necrosis” in man: analysis of 57 renal biopsies and a comparison with the glycerol model. Medicine 58: 362
    OpenUrlPubMed
  51. ↵
    Bialik, S., D. L. Geenen, I. E. Sasson, R. Cheng, J. W. Horner, S. M. Evans, E. M. Lord, C. J. Koch, R. N. Kitsis. 1997. Myocyte apoptosis during acute myocardial infarction in the mouse localizes to hypoxic regions but occurs independently of p53. J. Clin. Invest. 100: 1363
    OpenUrlCrossRefPubMed
  52. ↵
    Daemen, M. A., C. van’t Veer, G. Denecker, V. H. Heemskerk, T. G. Wolfs, M. Clauss, P. Vandenabeele, W. A. Buurman. 1999. Inhibition of apoptosis induced by ischemia-reperfusion prevents inflammation. J. Clin. Invest. 104: 541
    OpenUrlCrossRefPubMed
  53. ↵
    Goligorsky, M. S., E. Noiri. 1999. Duality of nitric oxide in acute renal injury. Semin. Nephrol. 19: 263
    OpenUrlPubMed
  54. ↵
    Markewitz, B. A., J. R. Michael, D. E. Kohan. 1993. Cytokine-induced expression of a nitric oxide synthase in rat renal tubule cells. J. Clin. Invest. 91: 2138
    OpenUrlCrossRefPubMed
  55. ↵
    Cohen, J., J. Carlet. 1996. INTERSEPT: an international, multicenter, placebo-controlled trial of monoclonal Ab to human tumor necrosis factor-α in patients with sepsis. Crit. Care Med. 24: 1431
    OpenUrlCrossRefPubMed
View Abstract
Back to top

In this issue

Print
Download PDF
Article Alerts
Email Article
Citation Tools

Jump to section