Introduction
Acute kidney injury (AKI) is an important contributor to morbidity and mortality among hospitalized patients. Sepsis is the leading cause of AKI in the critically ill patient, and currently no effective treatment exists [
1]. Septic AKI is generally believed to be due to regional hypoperfusion causing renal ischemia [
2]. However, recent experimental reports indicate that AKI may develop even though renal blood flow and blood pressure remain within physiological limits [
3]-[
5]. Thus, it is possible that the pathogenesis of septic AKI is complicated by factors other than ischemia. In line with this, we wanted to investigate whether an important receptor for initiating inflammation, the Toll-like receptor 4 (TLR4), participates in the pathophysiology of
Escherichia coli-induced AKI. TLRs have been identified as pivotal mediators in host defense as they are crucial in pathogen recognition and activation of the immune response [
6],[
7]. The main ligand for TLR4 is lipopolysaccharide (LPS), a component of the cell membrane of Gram-negative bacteria [
8]-[
10]. When LPS binds to TLR4, an inflammatory response is initiated via production and release of cytokines as well as stimulation of inflammatory cells [
11]. Targeting TLR4 with antibodies or specific inhibitors has shown beneficial effects by reducing mortality in experimental models of Gram-negative sepsis [
12]-[
17]. Initial clinical trials also showed beneficial effects of blocking TLR4 in sepsis [
18], but in larger follow-up investigations, survival was not improved [
19],[
20]. However, the renal effects of TLR4 antagonism are incompletely investigated, and there are reasons to believe that TLR4 may be an important mediator of sepsis-induced AKI. For example, the expression of TLR4 in renal tubules, glomeruli, and peri-tubular capillaries is increased after sepsis [
21], and mice deficient in TLR4 have a reduced increase in blood urea nitrogen (BUN) when subjected to LPS [
22]. Our group recently showed that pre-treatment with TAK-242 attenuated oliguria and reduced creatinine clearance in LPS-induced hyperdynamic shock [
3]. However, although the aforementioned results are promising, the potential of TLR4 modulation during sepsis and septic renal failure, as well as mechanisms of action, is still largely unknown.
One of the main objectives with the current study was to investigate whether TLR4 inhibition is effective in attenuating or reversing renal dysfunction even after sepsis has developed. Therefore, TAK-242 was administered 12 hours into normotensive ovine sepsis caused by an intravenous live
E. coli infusion, and the effect on renal function was followed for an additional 24 hours. To further explore the hypothesis that septic AKI is due to regional blood flow restriction, total renal blood flow and cortical and medullary microcirculation were continuously measured. Signs of ischemia were repeatedly monitored by microdialysis in the cortex and medulla. It is highly debatable whether septic AKI is associated with renal histopathology, and studies investigating structural damage to the kidney in long-term animal model of sepsis are lacking [
23]. Thus, both light and electron microscopy were used to analyze renal samples taken at the end of the 36-hour study period with regard to local injury and leukocyte infiltration.
Materials and methods
For detailed description of methods, please refer to the online supplemental material (Additional file
1). The experiments conform to the guidelines laid out in the Guide for the Care and Use of Laboratory Animals (National Academy of Science). The Regional Ethics Committee for Experiments in Animals, Stockholm, Sweden, approved the study in advance (N285/08).
Surgical preparation and study protocol
Twenty-seven adult Texel crossbred ewes were included in the study. Twenty-four of these sheep were anesthetized and prepared with catheters in the right carotid artery, the pulmonary artery, and the right jugular vein. For renal hemodynamic measurements, an ultrasonic flow probe was placed around the left renal artery, and two laser Doppler flow probes, one cortical and one medullar, were sutured on the left kidney. Intrarenal metabolism was studied by microdialysis catheters in the cortex and in the medulla, respectively. A urinary retention catheter was inserted into the bladder for urine sampling. After a post-surgical recovery period of 12 to 18 hours, the experiments commenced with the animals being conscious and placed in a pen. Sepsis was induced by an intravenous infusion of live E. coli bacteria (bolus of 3.9 × 109 colony-forming units followed by an infusion of 6.0 × 109/mL colony-forming units, starting at a rate of 0.2 mL/hour). The infusion rate was increased stepwise every 6 hours until reaching 4 mL/hour after 30 hours. After 12 hours of sepsis, sheep were randomized to receive a bolus dose (2 mg/kg) followed by a continuous infusion (4 mg/kg per 24 hours) of either the selective TLR4 inhibitor TAK-242 (10 mg/mL) (n = 7) or vehicle (n = 7). The treatment was blinded to the investigators, and the content of the infusions was revealed only after the experiments were performed. To exclude that surgery per se had a major impact on the results obtained in the TAK-242 and vehicle groups, an additional sheep served as time control. This included surgical preparation and recovery and monitoring for 36 hours but no E. coli infusion or treatment.
To investigate whether TAK-242 had any effect on renal function per se, an additional three sheep were surgically prepared and, in a cross-over design, subjected to treatment with either TAK-242 or vehicle. The treatment was initiated by a bolus dose (2 mg/kg) followed by a continuous infusion (2 mg/kg per 12 hours). No E. coli was administered.
Besides the intravenous fluids given post-surgery, fluid volume support was administered as Ringer's Acetate solution (B. Braun Melsungen, AG, Melsungen, Germany) intravenously at 1 mL/kg per hour and started 6 hours before the infusion of live E. coli bacteria. Blood samples (approximately 20 mL of venous blood and 1 mL of arterial blood) were drawn at baseline and every 6 hours after commencement of sepsis. Urinary output was measured and urine samples were collected every second hour. After 36 hours of sepsis, animals were deeply anaesthetized with sodium thiopental and terminated by an overdose of potassium chloride. The kidney was rapidly harvested and prepared for histological evaluation. The position of the laser Doppler probes and microdialysis catheters was confirmed visually by opening the kidney post-mortem. Renal biopsies were immediately frozen and stored at -70°C. If the animal was judged to be severely ill and in distress, it was euthanized prior to the end of the protocol.
Histological evaluation of renal biopsies
Small pieces of renal tissue were prepared and evaluated by light and transmission electron microscopy. In addition, some sections were stained with anti-myeloperoxidase antibody for easy quantification of polymorphonuclear leukocyte (PMN) cells in the glomeruli, the interstitium, and the peritubular capillaries.
Statistical analysis
Cardiovascular parameters were averaged off-line. Creatinine clearance was calculated as (urine flow × urine creatinine concentration)/plasma creatinine concentration. Cardiac output was indexed to body surface area (0.09 × body weight0.67).
All statistical calculations were performed by using Statistica 8.0 (Statsoft Inc., Tulsa, OK, USA), and the graphs were created with Sigma Plot 11.0 (SPSS Inc., Chicago, IL, USA). Data are expressed as means ± standard deviation of the mean or as mean and 95% confidence interval. Urine production, fractional sodium excretion (FENa), and urinary-N-acetyl-beta-D-glucosaminidase (U-NAG) values were transformed to follow a normal distribution by taking the logarithm of the raw data. Changes in parameters over time were analyzed according to a two-way repeated measures analysis of variance (ANOVA), with time as within effects and treatment (control/TAK-242) as between effects. If there was a significant interaction between time × treatment, an additional one-way repeated measures ANOVA was performed for each treatment to investigate whether that group changed significantly over time. The result of this analysis is not displayed in the figures but is referred to in the Results section. The significance level was set at a P value of not more than 0.05. Mann-Whitney U test was used to evaluate difference in PMN count and histological scores between treatments.
Discussion
In the present study, we aimed to investigate the renal effects of a TLR4-inhbitor, TAK-242, in experimental sepsis. The most striking finding was that TAK-242 reversed a progressive decline in renal function when administered therapeutically after 12 hours of hyperdynamic E. coli sepsis. This was associated with a prominent reduction in renal neutrophil infiltration, a decreased swelling of the endothelium in the glomerular capillaries, an attenuation of arterial and renal hyperlactemia, and less tubular damage as indicated by reduced U-NAG. Furthermore, no major effects on systemic or local renal hemodynamics by TAK-242 were seen during sepsis, and in separate experiments in animals without sepsis, TAK-242 had no obvious major effects per se.
Proper animal models of septic AKI are difficult to achieve, and too often they do not mimic the clinical scenario [
24]. In this study, we used fluid-resuscitated, conscious animals developing severe hyperdynamic sepsis caused by live bacteria. Detailed renal functional data were collected repeatedly over an extended time period and related to immune function and histological pathology. In addition, treatment was started when sepsis and renal dysfunction were already present. Another strength of this investigation was that the renal histological findings in large mimicked what was recently described in human kidney after sepsis [
6].
Innate immunity is the first line of defense against invading microbes and is crucial for preventing infections. TLR4 is central in this immune activation by binding LPS [
9],[
25], which results in an inflammatory response that induces the production and release of cytokines as well as stimulation of inflammatory cells [
11]. Among these cells are neutrophils, which are activated and stimulated to transmigrate from blood to tissue. This is caused by TLR4-mediated release of interleukin-1B (IL-1B), tumor necrosis factor-alpha (TNFα), and IL-6. Another important neutrophil activator released by TLR4 stimulation is serum amyloid protein 3, which mediates its effect by targeting the formyl peptide receptor [
26]. Although the inflammatory response is crucial for preventing infections, a strong activation of the innate immune system, like in sepsis, may inflict damage to endogenous tissue and impair organ function. TAK-242 binds selectively to the intracellular domain of TLR4 and disrupts the interaction with several adaptor molecules. This leads to inhibition of both the MyD88-dependent and the TRIF (TIR domain-containing adapter-inducing IFN-β)-dependent pathway for TLR4 signal transduction [
27].
Acute tubular necrosis (ATN) caused by ischemia has been assumed to be the underlying pathophysiology of septic AKI [
2],[
28]. Massive release of NO is believed to cause vasodilation that result in relative hypovolemia, a reduction in cardiac output, and hypotension. As a reflex, renal sympathetic nerve activity is increased together with elevated levels of vasopressin, endothelin, angiotensin II, and aldosterone. The resulting renal vascular vasoconstriction leads to ischemia and subsequent ATN that impairs renal function. Undoubtedly, ischemia is a crucial factor for many types of AKI. However, in sepsis, the situation may sometimes be different. Most patients with sepsis present with a hyperdynamic circulation with elevated cardiac output [
29]. Experimental data indicate that AKI develops although total renal blood flow is increased [
4],[
23] and hypotension is counteracted by vasoactive drugs [
3]. Furthermore, the correlation between ATN and septic renal failure is weak in both human and animal studies [
30]. The current results obtained in normotensive septic sheep confirm that hypotension is not a necessary attribute for septic AKI. This is in line with studies in critically ill patients with severe sepsis, in whom blood pressure levels did not correlate with the severity of renal failure [
31]. TAK-242 drastically improved renal function without major hemodynamic effects. Moreover, there was no sign of renal vasoconstriction, as total renal artery blood flow as well as cortical and medullary perfusion did not decline during the 36-hour experiment. Thus, renal hypoperfusion appears highly unlikely as a cause of septic AKI, given this experimental setting. A recently promoted hypothesis for sepsis-induced AKI is that the inflammatory response causes a preferential dilatation of renal efferent arterioles, increasing renal blood flow but reducing the hydrostatic pressure for glomerular filtration and thereby glomerular filtration rate (GFR) [
32]. The most important vasodilator released in sepsis is NO, and it has been suggested that excessive intra-renal NO could be responsible for the reduction in post-glomerular resistance [
33]. However, it is unlikely that the effect of TLR4 inhibition on renal function in this study is mediated by reduced NO formation as there were no differences in either MAP or renal blood flow and as TAK-242 did not affect NOx or cyclic guanosine monophosphate (cGMP) levels. This view is supported by results demonstrating no effect on renal function after intrarenal NO synthase inhibition during Gram-negative sepsis [
33].
Both cortical and medullary ratios of L/P were significantly increased by sepsis but subsequently reduced by TAK-242. In view of the preserved renal circulation, the renal hyperlactemia cannot easily be explained by reduced renal oxygen supply. However, direct measurement of renal tissue oxygenation or the utilization of oxygen by renal mitochondria was not performed in this study. Elevated lactate levels in sepsis not related to hypoxia have been linked to stimulation of muscle Na/K-ATPase [
34] and mitochondrial dysfunction [
35]. Thus, besides hypoxia, other factors may explain the renal hyperlactemia.
Sepsis caused severe endothelial swelling and decreased glomerular fenestration, both of which were reduced by TLR4 inhibition. The renal dysfunction observed may, to a large extent, be a consequence of decreased glomerular filtration per se, as creatinine clearance and the filtration fraction decreased significantly in the vehicle-treated animals, without significant tubular damage. Endothelial dysfunction in the peritubular capillaries has been highlighted as an important injury pathway in septic AKI [
36], possibly by inducing ATN [
37]. Herein, U-NAG did increase in vehicle-treated animals as a possible sign of tubular injury, but this took place several hours after renal function started to decline and was not confirmed by light microscopy. Instead, it is possible that endothelial swelling and perhaps decreased fenestration in the glomerulus may play an important role in the injury pathway and propagation of septic AKI. However, an early and specific marker of glomerular membrane dysfunction is urine protein leakage, and this was not detected in this study and is not a hallmark of sepsis-induced AKI.
Mobilization and recruitment of PMNs by the innate immune defense constitute a key event in response to an infection. After extravasation, PMNs destroy invading organisms by phagocytosis, release of acid hydrolases and antimicrobial peptides, and stimulation of antibiotic actions of monocytes and macrophages [
38]. However, PMN degranulation may also inflict damage to endogenous tissue. The role of PMNs in septic AKI is largely unknown, but after renal ischemia and reperfusion, PMNs have been shown to transmigrate from the circulation and contribute to AKI [
39]. In sepsis, Castoldi and colleagues [
40] recently showed that TLR4-deficient mice had reduced renal neutrophil activation and infiltration compared with wild-type mice and that neutrophil depletion improved renal function. This indicates that the extensive renal neutrophil accumulation caused by sepsis in this study is potentially deleterious. Activation of renal endothelial TLR4 has been suggested to play a critical role in the upregulation of adhesion molecules which can promote the recruitment of leukocytes to areas of injury and aggravate damage and inflammation in the tissue [
41],[
42]. A TLR4-dependent pathway promoting renal injury and inflammation in antibody-mediated glomeruli-nephritis and cisplatin-induced nephrotoxicity has also been described [
43],[
44]. Therefore, reduced TLR4-dependent infiltration of PMNs may have contributed to the improved renal function in the TAK-242 group. This is supported by recent findings of substantial renal leukocyte infiltration in human septic shock [
6]. Possible mechanisms for the renal dysfunction include PMN vascular endothelial attachment that reduces blood flow and causes ischemia [
45] or tubular damage by migrated PMNs [
46]. However, as discussed previously, no rheological effects of sepsis or TLR4 inhibition were noted in the current study, and light microscopy revealed no convincing evidence of general or widely spread structural damage to the tubules. Taken together, these observations indicate that neutrophil activation and recruitment into the kidney are potential and perhaps crucial mediators of septic AKI, although the underlying mechanism remains unknown. As shown in a series of elegant experiments by Watts and colleagues [
47]-[
49], TLR4 activation may also affect renal function by impairing tubular transport. In particular, HCO
3− reabsorption is inhibited by the direct effect of LPS on TLR4, and this may contribute to sepsis-induced acidosis.
Besides renal effects, TLR4 inhibition significantly attenuated the increase in mean pulmonary artery pressure and prevented the decrease in partial pressure of oxygen. The mechanism for these findings is unknown. However, a TLR4-dependent pathway for recruitment of neutrophils into the lung has been highlighted in endotoxemic mice [
41], and PMN degranulation may cause severe lung damage [
38]. The reason why arterial pCO
2 did not change is unknown but may be related to the fact that CO
2, compared with oxygen, more easily diffuses between blood and alveolus.
The current results are partly in contrast to the recent clinical trial using a TLR4 antagonist in sepsis [
19],[
20]. In that study, no effect on survival was discovered. However, data on renal function were not reported, and only a minority of the patients had verified Gram-negative sepsis. Thus, it is possible that some patients with sepsis would still benefit from TLR4 inhibition.
We would like to acknowledge some of the limitations of this study. Evaluation of the renal microcirculation was performed with the use of laser Doppler probes surgically implanted in both the cortex and medulla. This is a well-established method and is frequently used for assessment of renal tissue perfusion; however, the technique is invasive, and measurements are performed at a small proportion of the kidney, less than approximately 1 mm
3 at each probe site. Although the technique is a good option for continuous data acquisition [
50], it has problems detecting heterogeneity in microvascular flow. The use of creatinine clearance as an indicator for GFR is often used in the clinical setting, but the accuracy is limited in patients. However, in sheep, renal clearance of endogenous creatinine may be a relatively adequate measure of GFR [
51].
Acknowledgments
We are most grateful for the excellent laboratory assistance by Azar Baharpoor and Carina Nihlen, Department of Physiology and Pharmacology, Karolinska Institutet, Stockholm. Takeda Pharmaceutical Company Limited (Osaka, Japan) kindly provided us with TAK-242. The study was supported by funds from the Swedish Research Council (grant 521-2011-2843, RF), Karolinska University Hospital (JoF, JaF, AW, KH, AW, and VO), Karolinska Institutet (MR and SE), The Lars Hierta Foundation (JoF), The Swedish Society of Medicine (RF), The Swedish Society for Anaesthesia and Intensive Care (JoF), The Ruth and Richard Juhlin Foundation (RF), The Åke Wiberg Foundation (RF), and The Swedish Kidney Association (JoF). There was no involvement of the funding resources in research design collection, analysis, and interpretation of data. The writing and publication of the manuscript were done by the authors without any participation or influence from the funding sources for any of the authors.
Competing interests
The authors declare that they have no competing interests.