Background

Sepsis is a common cause of acute renal failure (ARF). The incidence of sepsis increases dramatically after 50 years of age; however, most ARF studies are performed in young mice.

Methods

We performed two common sepsis models, lipopolysaccharide (LPS) administration and cecal ligation puncture (CLP) in aged mice. We developed a fully treated CLP model in aged mice by treating mice with fluid resuscitation and antibiotics.

Results

LPS induced renal injury in aged but not young mice. However, volume resuscitation starting within 6 hours decreased renal injury. We then used this fluid resuscitation scheme, along with antibiotics, to develop a fully treated CLP model in aged mice. Mice subjected to CLP developed functional and histologic ARF and multiple organ damage. Treatment with ethyl pyruvate, even when started 12 hours after surgery, decreased serum creatinine, tubular damage, and multiple organ injury at 24 hours. Ethyl pyruvate decreased plasma tumor necrosis factor- (TNF-), and kidney mRNA for TNF, tissue factor, and plasminogen activator inhibitor-1 (PAI-1), and increased mRNA for urokinase-like plasminogen activator.

Conclusion

CLP in aged mice causes functional and histologic changes consistent with human ARF. A single dose of ethyl pyruvate inhibits renal and multiple organ damage, and is still effective when given 12 hours after surgery.

METHODS

Animals

Animal care followed National Institutes of Health (NIH) criteria for the care and use of laboratory animals in research. Young (7 to 8 weeks) and aged (42 to44 weeks) male C57BL/6 mice (NIH, Frederick, MD, USA) had free access to water and chow (NIH-07 Rodent Chow) (Zeigler Bros., Gardners, PA, USA) before and after surgery. Aged mice were housed individually.

LPS-induced ARF

A single dose of 0.08 mg/kg LPS (O11:B4) (Sigma Chemical Co., St. Louis, MO, USA) in 0.3 mL of sterile normal saline was administrated intravenously via the tail vein. In some animals, 1.5 mL of 3/4 normal saline was given at 0, 6, or 18 hours after LPS injection, and then every 12 hours. Animals were sacrificed at 48 hours for measurement of serum creatinine.

Polymicrobial CLP sepsis-induced ARF

Mice were anesthetized with 100 mg/kg ketamine, 10 mg/kg xylazine, and 1 mg/kg acepromazine intramuscularly. After laparotomy, a 5-0 silk ligature was placed 5 mm from the cecal tip. The cecum was punctured twice with a 21-gauge needle and gently squeezed to express a 1 mm column of fecal material. In sham-operated animals, the cecum was located, but neither ligated nor punctured. The abdominal incision was closed in two layers with 6-0 nylon sutures. After surgery, 1 mL of prewarmed normal saline was given intraperitoneally. All animals received a broad-spectrum antibiotic (imipenem/cilastatin; 25 mg/kg subcutaneously)³⁰ and 1.5 mL of 3/4 normal saline was administered at 6 and 18 hours after surgery by subcutaneous injection. Imipenem/cilastatin was selected because it has been used by several other laboratories in the young mouse CLP model and has been shown to increase survival more than triple antibiotic therapy^24,30,31. At the time of sacrifice, blood was collected from abdominal aorta for measurement of blood chemistries. The kidneys were harvested for histologic and mechanistic studies. In a very small pilot study, CLP surgery in young mice produced a wide range of creatinine values at 24 hours without much histologic evidence of renal damage (data not shown). Therefore, subsequent studies were performed in aged mice.

Survival study

Survival after surgery was assessed every 6 hours within the first 48 hours and then every 8 hours for 4 days. Antibiotic injection and fluid resuscitation were started 6 hours after CLP by subcutaneous injection, and then repeated every 12 hours for 4 days.

Ethyl pyruvate treatment protocol

Animals received 0.4 mL of Ringer's lactate (130 mmol/L Na⁺, 4 mmol/L K⁺, 2.7 mmol/L Ca²⁺, 109 mmol/L Cl^–, and 28 mmol/L lactate) or a similar volume of freshly made Ringer's ethyl pyruvate where ethyl pyruvate (Sigma Chemical Co.) was substituted for sodium lactate. A single dose was injected intraperitoneally at 0, 6, or 12 hours after CLP surgery. The dose of ethyl pyruvate was based upon that used by Ulloa et al²⁴.

Blood chemistries and serum TNF- measurements

Serum levels of BUN, aspartate transaminase (AST), alanine transaminase (ALT), amylase, creatine kinase (CK), and lactate dehydrogenase (LDH) were measured using an autoanalyzer (Hitachi 917, Boehringer Mannheim, Indianapolis, IN, USA). Serum creatinine was measured by a picric acid-based colorimetric kinetic autoanalyzer (Astra 8 autoanalyzer; Beckman Instruments, Fullerton, CA, USA). However, picric acid methods can greatly overestimate serum creatinine in mice³²; therefore we also measured serum creatinine by high-performance liquid chromatography (HPLC) adapted from Dunn and Sharma (Dr. Kumar Sharma, Thomas Jefferson University, personal communication, 2003), and Johns et al³³. Acetonitrile was added to serum, centrifuged, and the supernatant fraction was dried, resuspended in 5 mmol/L sodium acetate, pH 5.1, and chromatographed isocratically on a PRP-X200 cation exchange column (Hamilton, Reno NV, USA) and detected by ultraviolet absorbance at 234 nm (Agilent Technologies, Palo Alto, CA, USA). TNF- was assayed by enzyme-linked immunosorbent assay (ELISA) (QuantikineM, R&D Systems, Inc., Minneapolis, MN, USA). The minimum detectable concentration was less than 5 pg/mL.

Histologic examination

The 10% formalin-fixed, paraffin-embedded kidney sections were stained with periodic acid-Schiff reagent (PAS) or naphthol AS-D chloroacetate esterase (Sigma Chemical Co.). Histologic changes in the cortex and in the outer stripe of the outer medulla (OSOM) were assessed by quantitative measurements of tissue damage. Tubular damage was defined as tubular epithelial swelling, loss of brush border, vacuolar degeneration, necrotic tubules, cast formation, and desquamation. The degree of kidney damage was estimated at 400 magnification using five randomly selected fields for each animal by the following criteria: 0, normal; 1, areas of damage <25% of tubules; 2, damage involving 25% to 50% of tubules; 3, damage involving 50% to 75% of tubules; 4, 75% to 100% of the area being affected.

Western blot analysis for Cyr61

Cyr61 expression in the kidney was measured as described previously³⁴.

Measurement of mRNA abundance by reverse transcription-polymerase chain reaction

Reverse transcription-polymerase chain reaction (RT-PCR) was performed to examine malate dehydrogenase (MDH) (all primers shown 5' to 3') forward, GGTCATTGTTGTGGGAAACC; reverse, TCGACACGAACTCTCCCTCT; plasminogen activator inhibitor-1 (PAI-1) forward, CAGAGGTGGAAAGAGCCAGA; reverse, AGCGATGAACATGCTGAGG; tissue factor forward, GTGCAGGCATTCCAGAGAA; reverse, TGGGACAGAGAGGACCTTTG; tissue-type plasminogen activator (tPA) forward, TGATGGCTCAGAGCAACAAG; reverse, GCCAGGGTTGCACTAAACAT; urokinase-type plasminogen activator (uPA) forward, TTGTCCAAGAATGCATGGTG; reverse, GCTGCTCCACCTCAAACTTC; and TNF- forward, GAACTGGCAGAAGAGGCACT; reverse, AGGGTCTGGGCCATAGAACT) mRNA abundance. Briefly, total RNA was extracted from the kidney by using TRIzol (Invitrogen, Carlsbad, CA, USA), cDNA was generated SuperScript II (Invitrogen), and amplified by PCR primers. The PCR conditions were Mg 2.5 mmol/L, annealing temperature 57°C (59°C for uPA), and 35 cycles. These conditions resulted in a single band with a linear relationship between cDNA and PCR product. We normalized to mouse MDH in parallel PCR reactions. The PCR products were separated by electrophoresis and quantitated by NIH-Image.

Statistical analysis

All data are expressed as means SE. Differences between groups were examined for statistical significance by analysis of variance (ANOVA) with a multiple comparison correction (StatView 4.5, Berkeley, CA, USA; or SigmaStat 2.0, SPSS, San Rafael, CA, USA). A P value <0.05 was accepted as statistically significant.

RESULTS

Effect of volume resuscitation in a LPS-induced ARF model

Injection of 0.08 mg LPS caused a time-dependent transient increase in BUN and creatinine (measured by picric acid) in aged mice but not in young mice Figure 1a. Since the initial treatment of human sepsis is volume replacement, we restored volume with 1.5 mL of 3/4 normal saline every 12 hours, sufficient to prevent weight loss and keep plasma sodium constant. Volume repletion starting at the time of LPS injection ameliorated the increase in creatinine. Since the diagnosis of sepsis is often delayed, we determined the window of opportunity of volume repletion. Volume repletion starting at 6 hours but not 18 hours after LPS inhibited renal injury Figure 1a. Thus, aged mice are more susceptible to LPS-induced injury. However, the LPS-induced injury was unexpectedly reversed by fluid administration, which is not a feature of human sepsis-induced ARF⁸.

Figure 1.

Effect of volume resuscitation. (A) Volume replacement inhibits lipopolysaccharide (LPS)-induced acute renal failure (ARF) in aged mice. Aged mice were injected intravenously with 0.08 mg LPS. In some animals, volume replacement was started at 0, 6 or 18 hours, and continued every 12 hours. Animals were sacrificed at 48 hours for measurement of creatinine. Creatinine values are also shown for normal aged mice, sham-injected aged mice, and young mice given LPS but not fluids. Values are mean SE (N = 4 to 10 per group). *P < 0.05 vs. sham; +P < 0.05 vs. aged without fluid group. (B) Survival after antibiotic- and volume-resuscitated cecal ligation puncture (CLP) sepsis in aged mice. Aged mice were subjected to CLP. Antibiotic and fluid resuscitation were started at 6 hours after surgery, and then given every 12 hours. Survival rates were monitored for 96 hours. Closed circles and solid line indicates CLP group (N = 15). Open diamonds and dashed line indicates sham group (N = 10).

Full figure and legend (30K)

Characterization of the volume- and antibiotic-treated CLP model in aged mice

Since aged mice were more susceptible to LPS-induced ARF, we evaluated the CLP model in aged mice. To make the model more realistic, we administered both volume and antibiotic treatment every 12 hours starting at 6 hours after surgery to mimic the time needed to detect sepsis in humans. Under these conditions, body weight increased by 7%, and the survival was 100% at 24 hours, 43% at 48 hours, and 14% at 72 hours Figure 1b. Therefore, we set the end point of study at 24 hours to examine the mechanism of sepsis-induced ARF.

Aged mice given volume and antibiotic treatment after CLP had multiple organ injury Figure 2. AST, ALT, and LDH were increased as early as 6 hours after CLP, and increased further at 24 hours. The level of CK was also rapidly increased after CLP surgery; however, the transient increase at 1¹/₂ hours was also detected in the sham-surgery group and likely a result of the intramuscular anesthetic. CK increased above sham levels at 6 hours, and continued to rise at 24 hours. In contrast, release of serum amylase was delayed; amylase remained at basal levels during the first 12 hours and significantly increased at 24 hours after CLP. To confirm that these serologic tests reflected actual tissue damage, we performed histologic analysis in a parallel set of mice. We found moderately severe liver injury consisting of swollen and vacuolated hepatocytes with nuclear swelling, but no necrosis or inflammation. In the lung, we saw patchy congestion, hemorrhage, and inflammation (data not shown). This model reproduces the multiorgan damage following CLP in young mice and rats^30,35, and the hepatic, pancreatic, and muscle/myocardial damage in human sepsis^36,37,38.

Figure 2.

Time course of multiple organ injury. Antibiotic- and volume-resuscitated cecal ligation puncture (CLP) or sham surgery was performed as in Figure 1. Animals were sacrificed at the indicated time and blood samples were collected for measurement of serum aspartate transaminase (AST), alanine transaminase (ALT), amylase, creatine kinase (CK), and lactate dehydrogenase (LDH) (N = 5 to 6 per group). Closed circles and solid lines indicate CLP group. Open diamonds and dashed lines indicate sham group. *P < 0.05 vs. 0 hour.

Full figure and legend (29K)

CLP sepsis-induced ARF

CLP sepsis caused time-dependent increases in markers of renal dysfunction Figure 3. BUN was significantly increased as early as 3 hours after surgery, whereas the rise in creatinine was delayed until 12 and 24 hours after surgery. In contrast to previous nonfluid-replete CLP models where the BUN/creatinine ratio increases¹⁷, both BUN and creatinine increased in parallel in our fluid-replete model. This suggests a lack of significant volume depletion in this model. Studies by Myers et al³⁵ published in 1985 noticed that measurement of creatinine by picric acid overestimates serum creatinine concentration in mice. Therefore, we examined the levels of serum creatinine by HPLC technique using serum samples obtained from the same CLP mice. As shown Figure 3c, the level of HPLC creatinine in normal aged mice was 0.09 0.01 mg/dL, which was approximately one third that of picric acid creatinine. The increase in HPLC creatinine paralleled that of picric-acid creatinine; thus, we detected a significant increase in HPLC creatinine at 12 hours (0.24 0.02 mg/dL, P < 0.01 vs. 0 hour) and a further increase at 24 hours (0.37 0.07 mg/dL, P < 0.01 vs. 0 hour) after CLP. Creatinine was measured by HPLC in subsequent experiments.

Figure 3.

Time course of acute renal failure (ARF) after antibiotic- and volume-resuscitated cecal ligation puncture (CLP) sepsis in aged mice. Mice were treated as in Figure 2. Animals sacrificed at indicated times for measurement of serum creatinine by picric acid (A) or high-performance liquid chromatography (HPLC) method (C), and BUN (B). Data displayed as in Figure 2. *P < 0.05 vs. 0 hour.

Full figure and legend (25K)

Time course of tubular damage in CLP-ARF

At 6 hours after CLP, histologic examination of PAS-stained sections revealed focal tubular epithelial swelling, shortened brush border, and vacuolar degeneration in both the cortex and OSOM Figure 4b. At 12 to 24 hours after surgery, more extensive tubular damage was seen (Figure 4 c and f) in both areas. Unlike the ischemic or nephrotoxin-induced mouse ARF models^39,40, we did not detect tubular necrosis, vascular congestion, or excessive cast formation. Leukocyte accumulation, while occasionally present, was not a prominent feature Figure 4e.

Figure 4.

Time course of renal histologic changes. Kidney sections were stained with either periodic acid-Schiff (A to D) or naphthol AS-D chloroacetate esterase (E). Typical histology in sham-treated animals at 24 hours (A) or cecal ligation puncture (CLP) animals at 6 hours (B) and 24 hours (C) are shown. Typical histology in animals treated with ethyl pyruvate at 24 hours (D). Arrows, esterase-positive cells (original magnification, 250 (A to D), 400 (E). (F) Time course of the tubular damage. The tubular damage score (see Methods section) was measured in the cortex (right panel) or the outer stripe of the outer medulla (OSOM) (left panel). Values are mean SE (N = 5 to 6 per group). *P < 0.05 vs. sham. (G) Time course of renal Cyr61 expression. Renal Cyr61 measured at indicated times (hours) after CLP. N, normal; S, sham; and P, positive control for Cyr61.

Full figure and legend (136K)

Effect of CLP sepsis on renal Cyr61 expression and serum TNF-

We recently demonstrated that Cyr61 was rapidly induced in the kidney and secreted into the urine after ischemia/reperfusion injury, but not after volume depletion³⁴. Therefore, we measured Cyr61 to investigate the timing of tubular injury in polymicrobial sepsis-induced ARF. Cyr61 expression in the kidney was detected at 6 hours after surgery (a time at which serum creatinine was normal) Figure 3c, and sustained for at least 24 hours Figure 4g. Thus, renal injury occurs even before an increase in serum creatinine can be detected Figure 3c.

Both LPS and CLP sepsis induce a systemic inflammatory response, although the degree of activation is quite different¹⁰. We measured the levels of serum TNF- by ELISA Figure 5. Aged CLP mice had a biphasic increase in TNF-, with peaks at 3 and 24 hours. In contrast, aged mice injected with LPS had a more rapid, transient, and 20-fold larger increase in TNF- than detected in the CLP model. Thus, the proinflammatory stimulus in aged CLP mice is less intense than following LPS, similar to that reported in young mice¹⁰, but it also occurs much later.

Figure 5.

Time course of serum tumor necrosis factor- (TNF-) expression in lipopolysaccharide (LPS) and cecal ligation puncture (CLP) models in aged mice. Mice were treated as in Figure 2. At the indicated times, mice were sacrificed and the sera were collected (N = 5 to 6 per time point). Serum levels TNF- were measured by enzyme-linked immunosorbent assay (ELISA). Inset: Similar protocol in aged mice injected with LPS (N = 4 per group). Note dramatically different y-axis scales. *P < 0.05 vs. 0 hour.

Full figure and legend (15K)

Effects of ethyl pyruvate on CLP-induced multiple organ damage and ARF

Ulloa et al²⁴ reported that ethyl pyruvate greatly decreased mortality in an antibiotic- (but not fluid-) treated CLP model in young mice, even when started 24 hours after CLP. However, they did not measure renal function or indices of multiple organ damage. Therefore, we tested ethyl pyruvate in our volume-replete model of polymicrobial sepsis in aged mice; the same volume of Ringer's lactate was given as a control. As shown Figure 6, a single dose of either 8 (thin gray line) or 40 mg/kg (thick gray line) ethyl pyruvate immediately after surgery partially inhibited muscle (CK), hepatic, and pancreatic (amylase) injury at 24 hours. Delayed administration of ethyl pyruvate at 6 and 12 hours had similar protective effects; except that delay of ethyl pyruvate for 12 hours did not alter AST.

Figure 6.

Effects of ethyl pyruvate on multiple organ injury in antibiotic and volume resuscitated cecal ligation puncture (CLP) sepsis. Aged mice were treated as in Figure 2. A single dose of Ringer's lactate vehicle or 8 or 40 mg/kg ethyl pyruvate was injected at 0, 6, or 12 hoursr after CLP. Serum aspartate transaminase (AST), alanine transaminase (ALT), amylase, creatine kinase (CK), and lactate dehydrogenase (LDH) were measured (N = 10 to 13 per group). *P < 0.05 vs. sham. +P < 0.05 vs. Ringer's lactate.

Full figure and legend (39K)

The effects on renal function were more striking. A single dose of either 8 or 40 mg/kg of ethyl pyruvate after surgery significantly prevented the renal injury as measured by BUN or HPLC creatinine Figure 7 (8 mg/kg, 0.17 0.02 mg/dL; 40 mg/kg, 0.14 0.02 mg/dL vs. Ringer's lactate, 0.33 0.07 mg/dL). The higher dose of ethyl pyruvate significantly reduced the renal injury even when delayed until 6 or 12 hours after surgery (HPLC creatinine, 6 hours, 0.13 0.01 mg/dL; 12 hours, 0.17 0.03 mg/dL). Administration of 40 mg/kg ethyl pyruvate at either 0, 6, or 12 hours after CLP significantly reduced the tubular damage measured at 24 hours in both the cortex and OSOM Figure 4d and 7b.

Figure 7.

Effects of ethyl pyruvate on acute renal failure (ARF) following cecal ligation puncture (CLP) sepsis. Animals treated as in Figure 6, and were sacrificed at 24 hours (N = 10 to 13 per group). (A) Blood urea nitrogen (BUN) and high-performance liquid chromatography (HPLC) creatinine. (B) The tubular damage score was measured in the cortex and the outer stripe of the outer medulla (OSOM). *P < 0.05 vs. sham. +P < 0.05 vs. Ringer's lactate.

Full figure and legend (47K)

Effect of ethyl pyruvate on serum TNF-

To begin to investigate the mechanisms of action of ethyl pyruvate, we examined its effects on the systemic proinflammatory response to sepsis. The sepsis-induced increase in serum TNF- at 24 hours was inhibited by both 8 and 40 mg/kg of ethyl pyruvate administered at the time of surgery Figure 8a. Delay of ethyl pyruvate for either 6 or 12 hours also decreased serum TNF- at 24 hours Figure 8a.

Figure 8.

Effects of ethyl pyruvate on mediators after cecal ligation puncture (CLP) sepsis. Mice were subjected to antibiotic and volume resuscitated CLP sepsis. (A) Mice were treated as in Figure 6, and sacrificed at 24 hours. Tumor necrosis factor- (TNF-) measured by enzyme-linked immunosorbent assay (ELISA) (N = 5 to 6 per group). (B) A single dose of Ringer's lactate (RL) or 40 mg/kg ethyl pyruvate (EP) was injected after surgery, and mice were sacrificed at 12 hours to analyze the mRNA expression in the kidney using reverse-transcription-polymerase chain reaction (RT-PCR) analysis. (C) Semiquantitative analysis of RT-PCR (N = 6 per group). *P < 0.05 vs. sham. +P < 0.05 vs. Ringer's lactate. Abbreviations are: TF, tissue factor; PAI-1, plasminogen activator inhibitor-1; tPA, tissue-type plasminogen activator; uPA, urokinase-type plasminogen activator; MDH, malate dehydrogenase.

Full figure and legend (69K)

Effects of ethyl pyruvate on mRNA for kidney coagulation and fibrinolytic factors

There is important cross-talk between inflammation and coagulation pathway in sepsis. Activated protein C, which reduces mortality in human sepsis, modulates both inflammation and coagulation pathways^7,41. Therefore, we examined the expression mRNA for of key fibrinolytic coagulation factors in the kidney. As shown Figure 8b, CLP increased the expression of mRNA for TNF-, tissue factor, PAI-1, and tPA. In contrast, the mRNA expression of uPA was decreased at 12 hours. We found that a single 40 mg/kg dose of ethyl pyruvate injected after surgery inhibited the induction of mRNA for TNF-, tissue factor, PAI-1, and tPA, and also prevented the decrease in uPA mRNA abundance at 12 hours after surgery. Therefore, the ratio of PAI-1/uPA but not PAI-1/tPA was returned to basal levels by ethyl pyruvate at 12 hours after CLP Figure 8c. Taken together with the serum TNF- findings, our data suggest that ethyl pyruvate has effects on both proinflammatory and coagulation/fibrinolytic pathways.

DISCUSSION

We initially studied a LPS infusion model because of its simplicity; however, volume replacement with saline prevented and could also treat acute renal injury Figure 1a. Since human sepsis-induced ARF is not easily reversed with fluids⁸, we (1) developed a new mouse model of sepsis-induced ARF that mimics the human disorder, (2) used the model to demonstrate the effects of a novel intervention, ethyl pyruvate, to inhibit sepsis-induced ARF even when administered 12 hours after induction of sepsis, and (3) showed that ethyl pyruvate acts on a previously unsuspected coagulation/hemostatic pathway.

Model of sepsis-induced ARF that mimics the human disease

We developed a new model of sepsis-induced ARF based upon the polymicrobial CLP model, but with two distinctive features: (1) appropriate volume and antibiotic resuscitation, and (2) use of aged mice. As in the standard CLP model in young mice, animals were given an intra-abdominal infection and a small amount of fluid resuscitation immediately after surgery to foster a hyperdynamic phase that is characteristic of early human sepsis^14,42. Nevertheless, the animals may have been slightly volume depleted after surgery, since BUN was increased at 3 hours without a corresponding increase in creatinine Figure 3. Primate and porcine CLP models generally include fluid administration from the time of sepsis; continuous intravenous fluid replacement improves survival in a rat CLP model⁴³. However, the typical mouse CLP model does not include additional volume repletion and antibiotic treatment that is standard care for septic patients⁸. In our new model, the full volume and antibiotic resuscitation was not started until 6 hours after surgery to simulate a short window of undiagnosed sepsis. We used a broad-spectrum antibiotic regime employed by others³⁰. We chose a volume resuscitation schedule that inhibited LPS-induced renal damage Figure 1a and increased body weight by 7% to mimic the fluid overload often present in hypotensive intensive care unit patients. The second unique feature of the model was the use of aged mice. We used aged mice because they more closely match the age distribution of sepsis-induced ARF and because the elderly are more prone to sepsis-induced ARF². We also chose aged mice because we had difficulty obtaining reproducible renal injury in a small pilot study in young mice. Furthermore, aged mice are more prone to LPS-induced mortality⁴⁴, LPS-induced ARF Figure 1a, stress, and LPS-induced thrombosis, and glomerular fibrin deposition in part due to enhanced induction of NF-kB, TNF-, and PAI-1 mRNA induction in the kidney^45,46. Since this hyperresponsiveness to LPS was reduced in older PAI-1–deficient mice, some of the age effect is likely due to changes in PAI-1⁴⁵.

Despite adequate volume and antibiotic resuscitation, the aged animals developed progressive multiorgan injury first detected at 6 hours after sepsis, with initial changes in liver and muscle as previously described⁴⁷, and later pancreatic injury Figure 2. The kidney injury was delayed; whereas we detected a rise in the tubular injury marker Cyr61 at 6 hours, serum creatinine did not rise above sham-operated controls until 12 hours Figures 3c and 4g. The mortality rate was 60% at 48 hours and 86% at 72 hours, which is similar to the 75% mortality reported in human sepsis-induced ARF⁴. The temporal sequence of early muscle and liver injury, with delayed renal injury, reproduces the temporal sequence of human sepsis, where renal injury is often last in a series of organ failures.

A major new feature of this model is histologic evidence of ARF, which has been detected in only one nonfluid-replete CLP study²². We documented renal injury by three independent methods, including a new and more accurate measurement of serum creatinine by HPLC that detects smaller changes in creatinine from baseline than are possible with the picric acid–based method. First, we found parallel elevations in BUN and creatinine, suggestive of renal injury rather than volume depletion. Previous models have either failed to find increases in creatinine^20,21, or the change in BUN was much greater than that of creatinine, suggesting a component of volume depletion¹⁷. Second, we saw vacuolization of the proximal tubule throughout the cortex and OSOM. This vacuolated appearance is similar to the intracellular edema, loss of cell membrane, and mitochondrial swelling seen in the kidneys of young rats subjected to CLP without antibiotics or fluid replacement²². We did not detect necrosis of the proximal straight tubule, erythrocyte accumulation, or extensive cast formation typical of ischemic or nephrotoxic injury^39,40. The histologic appearance of human sepsis-induced ARF is poorly described. Biopsy is not routinely performed for the diagnosis of ARF, thus severely limiting the understanding of the process. Unfortunately, the kidney is subject to rapid autolysis, with changes that mimic those seen in ARF, which may limit their interpretation. Two "immediate" autopsy studies have been performed in patients who died of sepsis. Sato et al⁴⁸ detected only modest histologic and ultrastructural changes in immediate autopsy patients who died 4 to 50 days after onset of sepsis. They found increased vacuolization and flattening of the brush border membrane. A recent study by Hotchkiss et al⁴⁹ did not detect either necrosis or apoptosis in the kidneys of septic patients at autopsy; however, electron microcopy was not performed. Finally, we detected an increase in the renal tubular injury marker Cyr61 that we recently reported to increase in the kidney at 1 hour after renal ischemia/reperfusion but not after volume depletion³⁴. Thus, we detected functional, histologic, and surrogate marker evidence of renal injury. Taken together, this new model of volume and antibiotic resuscitated polymicrobial sepsis in aged mice shows many features in common with human sepsis, including sepsis-induced ARF.

It is interesting that we could detect subtle renal injury by increases in Cyr61 protein Figure 4g and renal pathology score in the outer stripe at 6 hours after injury Figure 4f, yet serum creatinine did not increase until 12 hours after injury Figure 3c. Dilution of creatinine at 6 hours by volume expansion, a frequent explanation for the inability to detect early renal injury using serum creatinine¹, is unlikely since the animals received fluid to replace surgical losses and the BUN was not similarly suppressed Figure 3b. Alternatively, renal cellular injury might occur before glomerular filtration rate (GFR) declines, suggesting that the tubules are injured before the microvasculature. In either case, better circulating or urinary markers are needed to detect early renal injury in the face of sepsis.

Effects of ethyl pyruvate on sepsis-induced ARF

Ethyl pyruvate has been shown to decrease mortality in both the LPS and CLP models and is effective even when given 24 hours after CLP sepsis²⁴. We have extended these results to show that ethyl pyruvate inhibits multiple organ injury in kidney, muscle, and pancreas Figures 6 and 7. Ethyl pyruvate still protected against renal and muscle injury even when treatment is started 12 hours after surgery; there was also a suggestion of pancreatic protection (not statistically significant). The renal injury was documented by suppression of sepsis-induced elevations of BUN, creatinine Figure 7a and tubular damage score, especially in the OSOM Figure 7b. The amount of protection was consistently 60% to 80% for HPLC and tubular injury score, without much change with delayed treatment. This wide therapeutic window, at least 12 hours for sepsis-induced ARF and 24 hours for sepsis-induced mortality²⁴, is longer than that of antibodies to triggering receptor expressed in myeloid cells (TREM) (4 hours)²³ or C5a (12 hours)³⁵, and similar to antibodies to HMG-B1 (24 hours)^27,50. Thus, ethyl pyruvate may be a treatment for both sepsis and sepsis-induced renal and multiorgan injury. This prolonged window of opportunity may be important clinically because of the difficulty in the early detection of sepsis and sepsis-induced ARF.

Mechanisms of action

The molecular targets and mechanisms of ethyl pyruvate are unknown. Whereas ethyl pyruvate was originally considered an ROS scavenger, recent data in hemorrhagic and LPS shock models suggest rather that ethyl pyruvate inhibits proinflammatory cytokines and HMG-B1. For example, ethyl pyruvate inhibits LPS-stimulated TNF- production from macrophages in vitro and plasma TNF- elevation after LPS injection in vivo, perhaps via inhibition of NF-kB and p38 MAPK signal transduction pathways^24,28,51. We found that ethyl pyruvate inhibited circulating TNF- measured at 24 hours Figure 8a, even when ethyl pyruvate was started 12 hours after surgery. If ethyl pyruvate was started 12 hours after CLP, the level of TNF- at 24 hours was similar to that obtained when ethyl pyruvate was administered 12 hours after CLP Figure 5. This suggests that delayed ethyl pyruvate treatment inhibited TNF- production even after the proinflammatory stimulus was underway. Ethyl pyruvate also modestly inhibited renal TNF- mRNA abundance, suggesting that ethyl pyruvate may act on nonrenal organs (liver, spleen, etc.), where most of the circulating TNF- is made in sepsis. However, the importance of this effect of ethyl pyruvate on TNF- to the overall pathogenesis of sepsis is uncertain. Anti-TNF- antibody therapy does not decrease mortality in CLP mice, and even increases overall mortality in human studies, perhaps because TNF- prevents the spread of a localized infection^10,52. In contrast, recent studies suggest that TNF- may be an important mediator of renal injury⁵³. Inhibition of TNF- by genetic or antibody techniques, or deletion of the TNF- receptor-1 inhibits renal injury following ischemia, cisplatin administration, and LPS-induced ARF^54,55,56,57. A subgroup analysis of the INTERSEPT study⁵⁸ found that anti-TNF- antibodies significantly decreased sepsis-induced ARF. Thus, TNF- may be an important mediator of renal damage. TNF- is usually synthesized at sites of injury and amplifies the inflammatory response; the resulting inflammation may injure surrounding tissues. However, renal inflammation is limited in the CLP model; neutrophils are present, but not prominent, in agreement with previous observations that renal myeloperoxidase is not elevated¹⁹ or only elevated twofold after CLP, despite large increases in liver and lung myeloperoxidase¹⁵. TNF- is more likely to act directly on the kidney. Kidney-specific deletion of the TNF receptor-1 decreases LPS-induced ARF⁵⁶, supporting a localized role for TNF- toxicity, possibly acting directly on tubular cells. Taken together, the studies suggest that ethyl pyruvate may protect against renal injury, in part, via its inhibition of TNF-.

Sepsis causes profound derangements in the coagulation/hemostatic system, including consumption of coagulation factors, increased coagulation/hemostatic activity, and reduced fibrinolysis²⁹. The positive result of a recent clinical trial of activated protein C⁷ suggests the pathophysiologic importance of the coagulation/hemostatic system in human sepsis. We found that sepsis decreased mRNA abundance for uPA and increased tissue factor and PAI-1 mRNA Figure 8b, which would tip the balance toward increased coagulation and reduced fibrinolysis, similar to previously studied effects of LPS⁵⁹. Ethyl pyruvate reversed these changes, perhaps restoring the renal hemostatic system to normal. Taken together, this suggests that ethyl pyruvate has a broader mechanism of action than previously suspected, with effects on proinflammatory cytokines and potential hemostatic pathways. Two other agents, activated protein C and c5a antibodies, that protect against sepsis also inhibit inflammatory and coagulation pathways^7,41,60.

CONCLUSION

We have introduced a novel volume- and antibiotic-treated aged mouse model of sepsis-induced ARF that mimics many aspects of human sepsis. This model should be useful for screening therapeutic agents and determining mechanisms of disease pathogenesis and protection. We then used the model to establish that ethyl pyruvate inhibits sepsis-induced renal and multiorgan damage, even when started 12 hours after surgery. Ethyl pyruvate protects against renal injury, in part, by inhibiting dysregulated inflammatory and coagulation/hemostatic pathways. Thus, it has the potential to simultaneously inhibit both early (TNF-, hemostatic) and late (HMG-B1) mediators in sepsis. Ethyl pyruvate may hold promise in clinical settings, since it can prevent the initiation and protect the progressive phase of sepsis-induced ARF.

References

References

1.	Star RA. Treatment of acute renal failure. Kidney Int 1998; 54: 1817−1831. | Article | PubMed | ISI | ChemPort |
2.	Neild GH. Multi-organ renal failure in the elderly. Int Urol Nephrol 2001; 32: 559−565. | Article | PubMed | ChemPort |
3.	Liaño F, Junco E & Pascual J et al. The spectrum of acute renal failure in the intensive care unit compared with that seen in other settings. Kidney Int 1998; 53 Suppl 66: S16−S24. | ISI |
4.	Neveu H, Kleinknecht D & Brivet F et al. 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 1996; 11: 293−299. | ISI | ChemPort |
5.	Angus DC, Linde-Zwirble WT & Lidicker J et al. Epidemiology of severe sepsis in the United States: Analysis of incidence, outcome, and associated costs of care. Crit Care Med 2001; 29: 1303−1310. | Article | PubMed | ISI | ChemPort |
6.	Brivet FG, Kleinknecht DJ, Loirat P & Landais PJ. Acute renal failure in intensive care units—Causes, outcome, and prognostic factors of hospital mortality; a prospective, multicenter study. French Study Group on Acute Renal Failure. Crit Care Med 1996; 24: 192−198. | Article | PubMed | ISI | ChemPort |
7.	Bernard GR, Vincent JL & Laterre PF et al. Efficacy and safety of recombinant human activated protein C for severe sepsis. N Engl J Med 2001; 344: 699−709. | Article | PubMed | ISI | ChemPort |
8.	De Vriese AS. Prevention and treatment of acute renal failure in sepsis. J Am Soc Nephrol 2003; 14: 792−805. | Article | PubMed | ISI |
9.	Esson ML & Schrier RW. Diagnosis and treatment of acute tubular necrosis. Ann Intern Med 2002; 137: 744−752. | PubMed | ISI |
10.	Remick DG, Newcomb DE, Bolgos GL & Call DR. Comparison of the mortality and inflammatory response of two models of sepsis: Lipopolysaccharide vs. cecal ligation and puncture. Shock 2000; 13: 110−116. | PubMed | ISI | ChemPort |
11.	Eskandari MK, Bolgos G & Miller C et al. Anti-tumor necrosis factor antibody therapy fails to prevent lethality after cecal ligation and puncture or endotoxemia. J Immunol 1992; 148: 2724−2730. | PubMed | ISI | ChemPort |
12.	Fink MP & Heard SO. Laboratory models of sepsis and septic shock. J Surg Res 1990; 49: 186−196. | Article | PubMed | ISI | ChemPort |
13.	Eichacker PQ, Parent C & Kalil A et al. Risk and the efficacy of antiinflammatory agents: Retrospective and confirmatory studies of sepsis. Am J Respir Crit Care Med 2002; 166: 1197−1205. | Article | PubMed | ISI |
14.	Wang P, Ba ZF & Chaudry IH. Increase in hepatic blood flow during early sepsis is due to increased portal blood flow. Am J Physiol 1991; 261: R1507−R1512. | PubMed | ISI | ChemPort |
15.	Matsukawa A, Kaplan MH & Hogaboam CM et al. Pivotal role of signal transducer and activator of transcription (Stat)4 and Stat6 in the innate immune response during sepsis. J Exp Med 2001; 193: 679−688. | Article | PubMed | ISI | ChemPort |
16.	Haybron DM, Townsend MC & Hampton WW et al. Alterations in renal perfusion and renal energy charge in murine peritonitis. Arch Surg 1987; 122: 328−331. | PubMed | ISI | ChemPort |
17.	Yang S & Hauptman JG. The efficacy of heparin and antithrombin III in fluid-resuscitated cecal ligation and puncture. Shock 1994; 2: 433−437. | PubMed | ISI | ChemPort |
18.	Matsukawa A, Hogaboam CM & Lukacs NW et al. Expression and contribution of endogenous IL-13 in an experimental model of sepsis. J Immunol 2000; 164: 2738−2744. | PubMed | ISI | ChemPort |
19.	Kalechman Y, Gafter U & Gal R et al. Anti-IL-10 therapeutic strategy using the immunomodulator AS101 in protecting mice from sepsis-induced death: Dependence on timing of immunomodulating intervention. J Immunol 2002; 169: 384−392. | PubMed | ISI | ChemPort |
20.	Pedersen PV, Warner BW & Bjornson HS et al. Hemodynamic and metabolic alterations during experimental sepsis in young and adult rats. Surg Gynecol Obstet 1989; 168: 148−156. | PubMed | ISI | ChemPort |
21.	Kuhlmann MK, Shahmir E & Maasarani E et al. New experimental model of acute renal failure and sepsis in rats. J Parenter Enteral Nutr 1994; 18: 477−485. | ISI | ChemPort |
22.	Huber-Lang M, Sarma VJ & Lu KT et al. Role of C5a in multiorgan failure during sepsis. J Immunol 2001; 166: 1193−1199. | PubMed | ISI | ChemPort |
23.	Bouchon A, Facchetti F, Weigand MA & Colonna M. TREM-1 amplifies inflammation and is a crucial mediator of septic shock. Nature 2001; 410: 1103−1107. | Article | PubMed | ISI | ChemPort |
24.	Ulloa L, Ochani M & Yang H et al. Ethyl pyruvate prevents lethality in mice with established lethal sepsis and systemic inflammation. Proc Natl Acad Sci USA 2002; 99: 12351−12356. | Article | PubMed | ChemPort |
25.	Sims CA, Wattanasirichaigoon S & Menconi MJ et al. Ringer's ethyl pyruvate solution ameliorates ischemia/reperfusion-induced intestinal mucosal injury in rats. Crit Care Med 2001; 29: 1513−1518. | PubMed | ISI | ChemPort |
26.	Tawadrous ZS, Delude RL & Fink MP. Resuscitation from hemorrhagic shock with Ringer's ethyl pyruvate solution improves survival and ameliorates intestinal mucosal hyperpermeability in rats. Shock 2002; 17: 473−477. | PubMed | ISI |
27.	Wang H, Bloom O & Zhang M et al. HMG-1 as a late mediator of endotoxin lethality in mice. Science 1999; 285: 248−251. | Article | PubMed | ISI | ChemPort |
28.	Yang R, Gallo DJ & Baust JJ et al. Ethyl pyruvate modulates inflammatory gene expression in mice subjected to hemorrhagic shock. Am J Physiol Gastrointest Liver Physiol 2002; 283: G212−G221. | PubMed | ISI | ChemPort |
29.	Hotchkiss RS & Karl IE. The pathophysiology and treatment of sepsis. N Engl J Med 2003; 348: 138−150. | Article | PubMed | ISI | ChemPort |
30.	Wang SC, Klein RD & Wahl WL et al. Tissue coexpression of LBP and CD14 mRNA in a mouse model of sepsis. J Surg Res 1998; 76: 67−73. | Article | PubMed | ISI | ChemPort |
31.	Newcomb D, Bolgos G, Green L & Remick DG. Antibiotic treatment influences outcome in murine sepsis: Mediators of increased morbidity. Shock 1998; 10: 110−117. | PubMed | ISI | ChemPort |
32.	Meyer MH, Meyer RA, Jr, Gray RW & Irwin RL. Picric acid methods greatly overestimate serum creatinine in mice: More accurate results with high-performance liquid chromatography. Anal Biochem 1985; 144: 285−290. | Article | PubMed | ISI | ChemPort |
33.	Johns BA, Broten T & Stranieri MT et al. Simple high-performance liquid chromatographic method to analyze serum creatinine has several advantages over the Jaffe picric acid reaction as demonstrated with a cimetidine dose response in rhesus monkeys. J Chromatogr B Biomed Sci Appl 2001; 759: 343−348. | Article | PubMed | ChemPort |
34.	Muramatsu Y, Tsujie M & Kohda Y et al. Early detection of cysteine rich protein 61 (CYR61, CCN1) in urine following renal ischemic reperfusion injury. Kidney Int 2002; 62: 1601−1610. | Article | PubMed | ISI | ChemPort |
35.	Huber-Lang MS, Sarma JV & McGuire SR et al. Protective effects of anti-C5a peptide antibodies in experimental sepsis. FASEB J 2001; 15: 568−570. | PubMed | ISI | ChemPort |
36.	Tribl B, Sibbald WJ & Vogelsang H et al. Exocrine pancreatic dysfunction in sepsis. Eur J Clin Invest 2003; 33: 239−243. | Article | PubMed | ISI | ChemPort |
37.	Betrosian A, Thireos E & Kofinas G et al. Bacterial sepsis-induced rhabdomyolysis. Intensive Care Med 1999; 25: 469−474. | Article | PubMed | ISI | ChemPort |
38.	Arlati S, Brenna S & Prencipe L et al. Myocardial necrosis in ICU patients with acute non-cardiac disease: a prospective study. Intensive Care Med 2000; 26: 31−37. | Article | PubMed | ISI | ChemPort |
39.	Chiao H, Kohda Y & McLeroy P et al. -Melanocyte-stimulating hormone protects against renal injury after ischemia in mice and rats. J Clin Invest 1997; 99: 1165−1172. | PubMed | ISI | ChemPort |
40.	Deng J, Kohda Y & Chiao H et al. Interleukin-10 inhibits ischemic and cisplatin-induced acute renal injury. Kidney Int 2001; 60: 2118−2128. | Article | PubMed | ISI | ChemPort |
41.	Joyce DE, Gelbert L & Ciaccia A et al. Gene expression profile of antithrombotic protein c defines new mechanisms modulating inflammation and apoptosis. J Biol Chem 2001; 276: 11199−11203. | Article | PubMed | ISI | ChemPort |
42.	Wang P, Zhou M & Rana MW et al. Differential alterations in microvascular perfusion in various organs during early and late sepsis. Am J Physiol 1992; 263: G38−G43. | PubMed | ISI | ChemPort |
43.	Smith EF, III, Slivjak MJ & Egan JW et al. Fluid resuscitation improves survival of endotoxemic or septicemic rats: Possible contribution of tumor necrosis factor. Pharmacology 1993; 46: 254−267. | PubMed | ISI | ChemPort |
44.	Tateda K, Matsumoto T, Miyazaki S & Yamaguchi K. Lipopolysaccharide-induced lethality and cytokine production in aged mice. Infect Immun 1996; 64: 769−774. | PubMed | ISI | ChemPort |
45.	Yamamoto K, Shimokawa T & Yi H et al. Aging accelerates endotoxin-induced thrombosis: Increased responses of plasminogen activator inhibitor-1 and lipopolysaccharide signaling with aging. Am J Pathol 2002; 161: 1805−1814. | PubMed | ISI | ChemPort |
46.	Yamamoto K, Takeshita K & Shimokawa T et al. Plasminogen activator inhibitor-1 is a major stress-regulated gene: Implications for stress-induced thrombosis in aged individuals. Proc Natl Acad Sci USA 2002; 99: 890−895. | Article | PubMed | ChemPort |
47.	Wang P, Ba ZF & Chaudry IH. Hepatocellular dysfunction occurs earlier than the onset of hyperdynamic circulation during sepsis. Shock 1995; 3: 21−26. | PubMed | ISI | ChemPort |
48.	Sato T, Kamiyama Y & Jones RT et al. Ultrastructural study on kidney cell injury following various types of shock in 26 immediate autopsy patients. Adv Shock Res 1978; 1: 55−69. | PubMed | ChemPort |
49.	Hotchkiss RS, Swanson PE & Freeman BD et al. Apoptotic cell death in patients with sepsis, shock, and multiple organ dysfunction. Crit Care Med 1999; 27: 1230−1251. | Article | PubMed | ISI | ChemPort |
50.	Czura CJ & Tracey KJ. Targeting high mobility group box 1 as a late-acting mediator of inflammation. Crit Care Med 2003; 31: S46−S50. | PubMed | ISI | ChemPort |
51.	Fink MP. Ethyl pyruvate: A novel anti-inflammatory agent. Crit Care Med 2003; 31: S51−S56. | PubMed | ISI | ChemPort |
52.	Echtenacher B, Weigl K, Lehn N & Mannel DN. Tumor necrosis factor-dependent adhesions as a major protective mechanism early in septic peritonitis in mice. Infect Immun 2001; 69: 3550−3555. | Article | PubMed | ISI | ChemPort |
53.	Schrier RW. Cancer therapy and renal injury. J Clin Invest 2002; 110: 743−745. | Article | PubMed | ISI | ChemPort |
54.	Daemen MARC, Van de Ven MWCM, Heineman E & Buurman WA. Involvement of endogenous interleukin-10 and tumor necrosis factor- in renal ischemia-reperfusion injury. Transplantation 1999; 67: 792−800. | PubMed | ISI | ChemPort |
55.	Knotek M, Rogachev B & Wang W et al. Endotoxemic renal failure in mice: Role of tumor necrosis factor independent of inducible nitric oxide synthase. Kidney Int 2001; 59: 2243−2249. | PubMed | ISI | ChemPort |
56.	Cunningham PN, Dyanov HM & Park P et al. Acute renal failure in endotoxemia is caused by TNF acting directly on TNF receptor-1 in kidney. J Immunol 2002; 168: 5817−5823. | PubMed | ISI | ChemPort |
57.	Ramesh G & Reeves WB. TNF-alpha mediates chemokine and cytokine expression and renal injury in cisplatin nephrotoxicity. J Clin Invest 2002; 110: 835−842. | Article | PubMed | ISI | ChemPort |
58.	Cohen J & Carlet J. INTERSEPT: An international, multicenter, placebo-controlled trial of monoclonal antibody to human tumor necrosis factor-alpha in patients with sepsis. International Sepsis Trial Study Group. Crit Care Med 1996; 24: 1431−1440. | Article | PubMed | ISI | ChemPort |
59.	Moll S, Schifferli JA & Huarte J et al. LPS induces major changes in the extracellular proteolytic balance in the murine kidney. Kidney Int 1994; 45: 500−508. | PubMed | ISI | ChemPort |
60.	Laudes IJ, Chu JC & Sikranth S et al. Anti-c5a ameliorates coagulation/fibrinolytic protein changes in a rat model of sepsis. Am J Pathol 2002; 160: 1867−1875. | PubMed | ISI | ChemPort |

Acknowledgments

This study was supported by the National Institutes of Heath. We thank Anthony Suffredini (NIH) for his comments and suggestions.