A two-hit mechanism for sepsis-induced impairment of renal tubule function

Bruns A. Watts 3rd, Thampi George, Edward R. Sherwood, David W. Good

Abstract

Renal insufficiency is a common and severe complication of sepsis, and the development of kidney dysfunction increases morbidity and mortality in septic patients. Sepsis is associated with a variety of defects in renal tubule function, but the underlying mechanisms are incompletely understood. We used a cecal ligation and puncture (CLP) model to examine mechanisms by which sepsis influences the transport function of the medullary thick ascending limb (MTAL). MTALs from sham and CLP mice were studied in vitro 18 h after surgery. The results show that sepsis impairs the ability of the MTAL to absorb HCO3 through two distinct mechanisms. First, sepsis induces an adaptive decrease in the intrinsic capacity of the tubules to absorb HCO3. This effect is associated with an increase in ERK phosphorylation in MTAL cells and is prevented by pretreatment of CLP mice with a MEK/ERK inhibitor. The CLP-induced reduction in intrinsic HCO3 absorption rate appears to involve loss of function of basolateral Na+/H+ exchange. Second, sepsis enhances the ability of LPS to inhibit HCO3 absorption, mediated through upregulation of Toll-like receptor 4 (TLR4)-ERK signaling in the basolateral membrane. The two inhibitory mechanisms are additive and thus can function in a two-hit capacity to impair renal tubule function in sepsis. Both effects depend on ERK and are eliminated by interventions that prevent ERK activation. Thus the TLR4 and ERK signaling pathways represent potential therapeutic targets to treat or prevent sepsis-induced renal tubule dysfunction.

  • thick ascending limb
  • lipopolysaccharide
  • TLR4
  • ERK
  • acid-base transport

sepsis is a major cause of morbidity and mortality in critically ill patients, accounting for more than 200,000 deaths per year in the United States alone and consuming considerable health care resources (9, 11, 42, 58). Renal insufficiency is a common and severe complication of sepsis, and the development of kidney dysfunction leads to prolonged hospitalization and doubles the risk for mortality in septic patients (11, 35, 42, 51, 54, 58). Bacterial sepsis and endotoxemia induce a variety of defects in renal tubule function in association with alterations in metabolic, fluid, and electrolyte homeostasis that contribute to sepsis pathogenesis (12, 35, 58). These include a urinary concentrating defect (30, 50), increased fractional excretion of sodium and glucose (50, 56, 57, 62), impaired glutamine metabolism (3), and hypotension (9, 57, 58), as well as the development of metabolic acidosis that contributes to multiple organ dysfunction (5, 9, 33, 36, 58) and is an independent risk factor for mortality in septic patients (17, 38, 49). Effective treatments for sepsis and sepsis-related kidney dysfunction are lacking, and there is a critical need to identify key mediators and molecular targets that may lead to new therapeutic strategies. At present, however, the pathophysiological mechanisms involved in renal tubule dysfunction during sepsis are poorly understood. To our knowledge, there have been no direct studies of the effects of sepsis on the transport function of renal tubule segments.

Recently, we demonstrated that molecules derived from Gram-negative and Gram-positive bacteria act directly through Toll-like receptors (TLRs) to impair the transport function of renal tubules, identifying a new pathophysiological mechanism that can contribute to renal tubule dysfunction during sepsis (24, 25). In particular, absorption of HCO3 by the medullary thick ascending limb (MTAL) is inhibited by lipopolysaccharide (LPS), the dominant cell wall molecule of Gram-negative bacteria (24). The direct action of LPS to inhibit renal tubule HCO3 absorption may exacerbate and/or impair the ability of the kidneys to correct systemic metabolic acidosis that contributes to sepsis pathogenesis (24, 25). Basolateral LPS inhibits HCO3 absorption in the MTAL through activation of its cell-surface receptor TLR4, which results in the downstream activation of an ERK-dependent signaling pathway coupled to inhibition of the apical membrane NHE3 Na+/H+ exchanger (24, 66). Previous studies have shown that TLR4 expression is increased in nephron segments of septic rats (13) and that sepsis increases ERK activity in cells of the lung and liver (72). Whether sepsis alters innate immune signaling pathways or intracellular signaling molecules that influence renal tubule transport processes has not been directly investigated. Understanding the molecular events by which sepsis impairs renal tubule transport would provide insight into how bacterial molecules adversely affect the function of epithelial cells and could aid in identifying potential therapeutic strategies to attenuate renal tubule dysfunction during sepsis.

The purpose of the present study was to examine cellular and molecular mechanisms by which sepsis impairs renal tubule function. The effects of sepsis on HCO3 absorption by the MTAL were investigated using a cecal ligation and puncture (CLP) model in the mouse. The CLP model has been used extensively to study sepsis pathogenesis because it reproduces important clinical features of human sepsis (11). The results show that CLP-induced sepsis impairs HCO3 absorption in the MTAL through two distinct mechanisms: 1) by decreasing the intrinsic capacity of the MTAL to absorb HCO3, and 2) by enhancing LPS-induced inhibition of HCO3 absorption, mediated through upregulation of basolateral membrane TLR4 signaling. Both mechanisms were found to depend on ERK activation, identifying the ERK pathway as a potential therapeutic target for treatment or prevention of sepsis-induced renal tubule dysfunction.

MATERIALS AND METHODS

Animals.

C57BL/6J mice 8 to 11 wk old were purchased from The Jackson Laboratory (Bar Harbor, ME). The animals were maintained under pathogen-free conditions in microisolator cages and had free access to standard rodent chow and water throughout the experiments. All experimental procedures were performed in accordance with criteria established by the National Institutes of Health Guide for the Care and Use of Laboratory Animals and were approved by the Institutional Animal Care and Use Committee of the University of Texas Medical Branch.

Cecal ligation and puncture.

CLP was performed as previously described (47, 53, 60). Mice were anesthetized with 2% isoflurane and a 1- to 2-cm midline incision was made through the abdominal wall. The cecum was identified and ligated 1 cm from the tip with a silk tie. A double puncture of the cecal wall was performed using a 20-gauge needle, and cecal contents were expressed from the puncture site to ensure a full-thickness perforation. Care was taken not to obstruct flow between the ileum and colon. The cecum was returned to the abdominal cavity, and the incision was closed with surgiclips followed by intraperitoneal administration of 1 ml of prewarmed isotonic saline for fluid resuscitation. Mice received buprenorphine (0.05–0.1 mg/kg) subcutaneously at the time of surgery. Sham mice underwent an identical procedure except that the cecum was neither ligated nor punctured. In some experiments (see Figs. 4 and 10), mice were pretreated with the MEK1/2 inhibitor PD98059 (16 mg/kg in DMSO/isotonic saline 1:25, 250 μl ip) 1 h before surgery. At 18 h after surgery, mice were anesthetized and kidneys were removed for isolated tubule studies as previously described (24, 29). Arterial blood was collected in heparinized syringes from the carotid artery of identically treated animals. This model has been characterized previously and reproduces key features of human sepsis, including a hyperdynamic circulation, metabolic acidosis, elevated serum levels of proinflammatory cytokines, late-phase immunosuppression, and multiple organ involvement including kidney dysfunction (10, 11, 40, 46, 47, 53, 55, 60; see results). Mice were studied 18 h after surgery because this time point was found to induce reproducible changes in MTAL function (see results) and has been used in other studies to assess CLP-induced effects on the kidney and immune system (10, 39, 46, 53, 55, 60).

Body temperature was measured using a rectal thermometer. Serum creatinine concentration was measured in the University of Texas Medical Branch Clinical Chemistry Laboratory using a creatinine amidinohydrolase/sarcosine oxidase-based enzymatic assay. Cytokine levels (TNF-α and IL-6) in plasma and whole kidney homogenates were measured using an ELISA according to the manufacturer's protocol (eBioscience). Cytokine concentrations were determined by measuring optical density at 450 nm using a microtiter plate reader (Dynatech Laboratories).

Tubule perfusion and measurement of net HCO3 absorption.

MTALs were isolated and perfused in vitro as previously described (18, 24). Tubules were dissected from the inner stripe of the outer medulla at 10°C in bath solution (see below), transferred to a bath chamber on the stage of an inverted microscope, and mounted on concentric glass pipettes for perfusion at 37°C. The tubules were perfused and bathed under basal conditions in a solution that contained the following (in mM): 146 Na+, 4 K+, 122 Cl, 25 HCO3, 2.0 Ca2+, 1.5 Mg2+, 2.0 phosphate, 1.2 SO42−, 1.0 citrate, 2.0 lactate, and 5.5 glucose (equilibrated with 95% O2-5% CO2, pH 7.45 at 37°C). Solutions containing LPS (ultra pure Escherichia coli K12; InvivoGen) and other experimental agents were prepared as described previously (24, 29, 65, 66). Experimental agents were added to the bath solution as described in results. Transepithelial voltage was measured between calomel cells using 140-mM NaCl-agar bridges as previously described (19). In each experimental series, MTALs from CLP mice were studied concurrently with tubules from sham mice obtained in the same shipment.

The protocol for study of transepithelial HCO3 absorption was as described previously (18, 24, 65, 66). Tubules were equilibrated for 20–30 min at 37°C in the initial perfusion and bath solutions, and the luminal flow rate (normalized per unit tubule length) was adjusted to 1.5–1.9 nl·min−1·mm−1. One to three 10-min tubule fluid samples were then collected for each period (initial, experimental, and recovery). The tubules were allowed to reequilibrate for 5–10 min after an experimental agent was added to or removed from the bath solution. The absolute rate of HCO3 absorption (JHCO3, pmol·min−1·mm−1) calculated from the luminal flow rate and the difference between total CO2 concentrations was measured in perfused and collected fluids (18). An average HCO3 absorption rate was calculated for each period studied in a given tubule. When repeat measurements were made at the beginning and end of an experiment (initial and recovery periods), the values were averaged. Single tubule values are presented in results (see Figs. 2, 4, 5, 6, 7, 10, and 11). Mean values ± SE (n = number of tubules) are presented in the text. The absolute decrease in HCO3 absorption was calculated for individual tubules as the difference between absorption rates measured in the absence and presence of experimental agent (bath LPS). The fractional decrease in HCO3 absorption is the absolute decrease expressed as a percentage of the basal absorption rate measured in the same tubule.

Immunoblotting.

Immunoblot analysis of phosphorylated ERK1/2 was carried out on the inner stripe of the outer medulla as previously described (23, 66). This tissue preparation is highly enriched in MTALs and exhibits regulatory changes in signaling proteins that accurately reflect changes observed in the MTAL (23, 65, 66, 68). Tissue samples were homogenized in ice-cold PBS and lysed for 2 h at 4°C in RIPA buffer with protease inhibitors. Samples of equal protein content (50 μg/lane) were separated by SDS-PAGE using 9% gels and transferred to PVDF membranes. Membranes were blocked with 5% BSA in TBS-Tween and incubated overnight at 4°C with anti-phospho-ERK1/2-Thr202/Tyr204 (1:2,500) or anti-ERK1/2 (1:5,000) antibody (Cell Signaling Technology). After being washed in TBS, horseradish peroxidase-conjugated anti-rabbit secondary antibody was applied and immunoreactive bands were detected by chemiluminescence (luminol reagent; Santa Cruz Biotechnology). Protein bands were quantified by densitometry.

Confocal immunofluorescence microscopy.

MTALs were studied by confocal microscopy as previously described (23, 64, 66). MTALs were dissected and mounted on Cell-Tak-coated coverslips at 10°C. The tubules were then incubated for 15 min at 37°C in a flowing bath using the same solution as in HCO3 transport experiments. In one series of experiments (see Fig. 8), tubules were incubated in the absence and presence of LPS. Following incubation, the tubules were washed with PBS and fixed and permeabilized in acetone at −20°C for 10 min. The tubules were incubated in Image-iT FX signal enhancer (Invitrogen) for 30 min at room temperature, washed, and blocked in 10% goat serum in PBS for 1 h at room temperature. The tubules were then incubated overnight at 4°C with anti-phospho-ERK1/2 (1:200; Cell Signaling Technology) or anti-TLR4 (1:100; Abcam) antibody. They were washed and then incubated for 1 h at room temperature in Alexa 488-conjugated goat anti-rabbit IgG antibody (1:100; Invitrogen) in blocking buffer. Fluorescence staining was examined using a Zeiss laser-scanning confocal microscope (LSM510 UV META), as described previously (23, 64, 66). Tubules were imaged longitudinally and z-axis optical sections (0.4 μm) were obtained through a plane at the center of the tubule, which provides a cross-sectional view of cells in the lateral tubule walls. For individual experiments, two to four tubules from the same kidney for each experimental condition, or from sham and CLP mice, were fixed and stained identically and imaged in a single session at identical settings of illumination, gain, and exposure time. The fluorescence intensity of p-ERK1/2 and TLR4 staining was quantified as previously described (23, 27, 66). Two-dimensional image analysis was performed using MetaMorph software in which boxes were positioned on the cytoplasm in the mid-region of the cell (1.4 × 4.2 μm; p-ERK) or on linear regions of the basolateral membrane domain (0.7 × 1.2 μm; TLR4), and pixel intensity per unit area was determined for each region. Three different cells were analyzed per optical section, and three optical sections were analyzed per tubule, one section at the center of the tubule and two sections positioned 0.12 μm above and below the center section. The measurements were averaged to obtain a value for each tubule. The fluorescence intensity for experimental groups was expressed as a percentage of the control value measured in the same experiment. Mean values (n = number of tubules) were used for statistical analysis. Morphometric analysis showed no difference in total cell volume or cell surface area in MTALs from sham and CLP mice.

Statistical analysis.

Results are presented as means ± SE. Differences between means were evaluated using Student's t-test for paired or unpaired data, as appropriate. P < 0.05 was considered statistically significant.

RESULTS

Systemic features of CLP model.

Mice were studied at 18 h after sham or CLP surgery, a time point at which CLP was found to induce reproducible changes in MTAL function (see below). The survival rate was 100% for both sham and CLP mice over the 18-h interval (in separate survival studies the mortality rate for CLP mice was 80% at 36 h and 100% at 60 h; no mortality was observed in sham mice at any time point). Body weight did not differ in sham and CLP mice (17.2 ± 0.4 g, sham vs. 17.4 ± 0.3 g, CLP; P = NS) and did not change over the 18-h postsurgery period in either group, indicating adequate volume replacement (10). Systemic measurements at 18 h after surgery showed that CLP decreased body temperature, increased plasma cytokine concentrations, decreased plasma HCO3 concentration, increased serum creatinine concentration, and increased kidney cytokine levels (Fig. 1). These results are similar to findings reported previously in CLP mice (10, 39, 46, 53, 55, 60) and, together with the MTAL data presented below, provide evidence of involvement of the kidneys in the sepsis response.

Fig. 1.

Systemic effects of cecal ligation and puncture (CLP). Mice underwent sham or CLP procedures as described in materials and methods and body temperature (A), plasma TNF-α and IL-6 concentrations (B), plasma HCO3 concentration (C), serum creatinine concentration (D), and kidney TNF-α and IL-6 concentrations (E) were measured at 18 h after surgery. *P < 0.05 vs. sham (n = 5 to 10 mice per group). Cytokine concentrations in kidneys of sham mice were below the limit of detection (8 pg/ml for TNF-α and 4 pg/ml for IL-6).

CLP decreases basal HCO3 absorption in the MTAL.

HCO3 absorption rates were determined in isolated, perfused MTALs from sham and CLP mice 18 h after surgery. The HCO3 absorption rate was decreased by 22% (from 15.2 ± 0.2 to 11.8 ± 0.3 pmol·min−1·mm−1; P < 0.001) in MTALs from the CLP mice (Fig. 2A). CLP had no effect on transepithelial voltage, an indirect measure of NaCl absorption rate in the mouse MTAL (Fig. 2B). Thus the decrease in HCO3 absorption is not the result of a nonspecific metabolic or cytotoxic effect of sepsis on the MTAL cells. The rate of HCO3 absorption in MTALs from sham-operated mice did not differ from that measured concurrently in tubules from control mice that did not undergo surgery (15.1 ± 0.4 pmol·min−1·mm−1; n = 8). These results demonstrate that CLP-induced sepsis decreases the capacity of the MTAL to absorb HCO3.

Fig. 2.

Effects of CLP on HCO3 absorption (A) and transepithelial voltage (B; VTE) in the medullary thick ascending limb (MTAL). MTALs from sham and CLP mice were isolated and perfused in vitro 18 h after surgery. Data points are average values for single tubules. JHCO3, absolute rate of HCO3 absorption. P values compare sham vs. CLP (unpaired t-test). NS, not significant. Number of tubules is 8 for sham and 12 for CLP. Mean values are given in results.

CLP increases ERK phosphorylation in the MTAL.

We have shown previously that activation of ERK decreases HCO3 absorption in the MTAL (23, 65, 66, 68). To determine whether CLP may influence HCO3 absorptive capacity in the MTAL through the ERK pathway, we examined the effect of CLP on ERK phosphorylation. MTALs dissected from sham and CLP mice 18 h after surgery were stained with anti-phospho-ERK1/2 (p-ERK) antibody and then analyzed by confocal immunofluorescence (Fig. 3, A and B). CLP increased p-ERK staining 1.4 ± 0.1-fold. The effect of CLP on ERK phosphorylation was examined further by immunoblot analysis of the inner stripe of the outer medulla, the region of the kidney highly enriched in MTALs. As shown in Fig. 3, C and D, CLP increased ERK phosphorylation 1.8 ± 0.1-fold without a change in total ERK level. These results show that the effect of CLP to reduce HCO3 absorptive capacity in the MTAL is associated with activation of the ERK pathway.

Fig. 3.

Effect of CLP on ERK phosphorylation in the MTAL. A: MTALs dissected from sham and CLP mice 18 h after surgery were fixed and stained with anti-phospho-ERK1/2-Thr202/Tyr204 (p-ERK) antibody. Tubules were analyzed by confocal immunofluorescence as described in materials and methods. Images are z-axis sections (0.4 μm) taken through a plane at the center of the tubule showing a cross-sectional view of cells in the lateral tubule walls. CLP increased p-ERK labeling. Images are representative of at least 6 tubules of each type. Scale bar = 5 μm. B: intensity of p-ERK staining for experiments in A was quantified as described in materials and methods. CLP value is presented as a percentage of the sham level measured in the same experiment. Bars are means ± SE. *P < 0.05 vs. sham. C: inner stripe tissue from kidneys of sham and CLP mice was obtained 18 h after surgery and cell lysates were immunoblotted with p-ERK antibody for analysis of ERK phosphorylation and anti-ERK antibody for total ERK level. Blots show 3 independent experiments. D: p-ERK level normalized for total ERK was determined for experiments in C by densitometry. p-ERK-to-ERK ratio for CLP mice is presented as a percentage of the sham value. Bars are means ± SE. *P < 0.05 vs. sham.

An inhibitor of ERK activation prevents the CLP-induced decrease in basal HCO3 absorption.

To test whether ERK activation may play a role in the reduced HCO3 absorptive capacity in MTALs from CLP mice, mice were pretreated with the MEK inhibitor PD98059 (16 mg/kg ip) 1 h before surgery. This treatment protocol has been shown to suppress CLP-induced ERK phosphorylation in lung and liver (72). MTALs were isolated from PD98059-treated mice 18 h after sham or CLP surgery, and the HCO3 absorption rate was determined. As shown in Fig. 4A, CLP had no effect on basal HCO3 absorption rate in MTALs from PD98059-treated mice (15.2 ± 0.6 pmol·min−1·mm−1, PD98059 + sham vs. 15.2 ± 0.4 pmol·min−1·mm−1, PD98059 + CLP; P = NS). The rate of HCO3 absorption in MTALs from PD98059-treated sham mice was similar to that measured in tubules from sham mice not given the inhibitor (Fig. 2A). These results show that the effect of CLP to reduce HCO3 absorptive capacity in the MTAL is prevented by pretreatment with an inhibitor of ERK activation.

Fig. 4.

Pretreatment with PD98059 prevents CLP-induced effects on HCO3 absorption and ERK phosphorylation in the MTAL. Mice were given PD98059 (16 mg/kg ip) 1 h before sham or CLP surgery. A: MTALs from PD98059-treated mice were isolated and perfused in vitro 18 h after surgery for measurement of HCO3 absorption rate. JHCO3, data points, and P value are as in Fig. 2A. Number of tubules is 6 for PD98059 + sham and 5 for PD98059 + CLP. Mean values are given in results. B: MTALs from PD98059-treated mice were isolated 18 h after surgery and analyzed for p-ERK level by confocal immunofluorescence as in Fig. 3, A and B. Images are representative of at least 9 tubules of each type. C: intensity of p-ERK staining for experiments in B was quantified as described in materials and methods. PD98059 + CLP value is presented as a percentage of the PD98059 + sham level measured in the same experiment. Bars are means ± SE. PD98059 + CLP value does not differ from PD98059 + sham.

Further studies were carried out to examine the effect of PD98059 pretreatment on ERK activation. MTALs from mice pretreated with PD98059 were isolated 18 h after sham or CLP surgery and analyzed for ERK phosphorylation by confocal immunofluorescence. As shown in Fig. 4, B and C, pretreatment with PD98059 prevented the increase in ERK phosphorylation induced in the MTAL by CLP. The effect of CLP to increase ERK phosphorylation in the MTAL was not affected by pretreatment with vehicle alone (not shown). These results show that the ability of PD98059 pretreatment to restore the basal HCO3 absorption rate to normal values in MTALs from septic mice correlates directly with suppression of CLP-induced ERK activation.

Inhibition of HCO3 absorption by bath amiloride is eliminated in MTALs from CLP mice.

We have shown previously that activation of ERK inhibits HCO3 absorption in the MTAL through inhibition of the basolateral NHE1 Na+/H+ exchanger (23, 29, 63, 68). To test whether this mechanism may be involved in mediating the CLP-induced decrease in the basal HCO3 absorption rate, we examined the effect on HCO3 absorption of 10 μM bath amiloride, which inhibits HCO3 absorption in the MTAL through inhibition of basolateral NHE1 (21, 29, 63, 64). In MTALs from sham mice, adding amiloride to the bath decreased HCO3 absorption by 27%, from 14.6 ± 0.2 to 10.6 ± 0.5 pmol·min−1·mm−1 (Fig. 5A). In contrast, adding amiloride to the bath had no effect on HCO3 absorption in MTALs from CLP mice (11.6 ± 0.2 pmol·min−1·mm−1, basal vs. 11.6 ± 0.1 pmol·min−1·mm−1, bath amiloride; Fig. 5B). These results suggest that CLP and bath amiloride inhibit HCO3 absorption in the MTAL through a common mechanism and that CLP may reduce basal HCO3 absorptive capacity through an NHE1-dependent pathway.

Fig. 5.

Effect of bath amiloride on HCO3 absorption in the MTAL. MTALs from sham (A) and CLP (B) mice were perfused in vitro under basal conditions, and then amiloride (10 μM) was added to and removed from the bath solution. MTALs were studied 18 h after surgery. Data points and JHCO3 are as in Fig. 2. Lines connect paired measurements made in the same tubule. P values are for paired t-test. Mean values are given in results.

CLP enhances inhibition of HCO3 absorption by basolateral LPS in the MTAL.

Previously, we demonstrated that LPS inhibits HCO3 absorption in the MTAL through TLR4 (24, 66). To test whether the ability of LPS to inhibit HCO3 absorption is influenced by sepsis, we examined the effect of basolateral LPS on HCO3 absorption in MTALs from sham and CLP mice. Tubules were perfused in vitro 18 h after surgery. Adding LPS to the bath decreased HCO3 absorption by 26% (from 15.5 ± 0.5 to 11.5 ± 0.5 pmol·min−1·mm−1; P < 0.001) in MTALs from sham mice compared with a decrease of 44% (from 11.4 ± 0.6 to 6.4 ± 0.6 pmol·min−1·mm−1; P < 0.001) in MTALs from CLP mice (Fig. 6A). Both the fractional and absolute decreases in HCO3 absorption induced by bath LPS were significantly higher in tubules from the CLP mice (Fig. 6B). The inhibition by LPS in MTALs from sham mice is similar to that shown previously in tubules from mice not undergoing surgery (27, 66). These results show that the ability of basolateral LPS to inhibit HCO3 absorption is enhanced in MTALs from septic mice. In addition, the effect of CLP to enhance inhibition by LPS is additive to the CLP-induced reduction in basal HCO3 absorption rate, such that the combination of the two effects reduces the HCO3 absorption rate by ∼ 60% compared with control tubules not exposed to bacterial stimuli.

Fig. 6.

CLP enhances inhibition of HCO3 absorption by basolateral LPS in the MTAL. A: MTALs from sham and CLP mice were perfused in vitro under basal conditions, and then LPS (500 ng/ml) was added to and removed from the bath solution. Tubules were studied at 18 h after surgery. JHCO3, data points, lines, and P values are as in Fig. 5. Mean values are given in results. B: fractional and absolute decreases in HCO3 absorption induced by bath LPS in experiments in A. Bars are means ± SE. *P < 0.05 vs. sham.

Basolateral LPS inhibits HCO3 absorption through an ERK-dependent pathway in MTALs from sham and CLP mice.

Previously, we demonstrated that basolateral LPS inhibits HCO3 absorption in the MTAL through activation of the ERK pathway (66). To test whether the increased ability of LPS to inhibit HCO3 absorption in MTALs from CLP mice depends on ERK, we examined the effects of LPS in tubules bathed with PD98059 or U0126, two selective inhibitors of MEK1/2 that block activation of ERK1/2 in the MTAL (23, 65, 68). As shown in Fig. 7, the inhibition of HCO3 absorption by bath LPS was eliminated completely by the MEK/ERK inhibitors in MTALs from both sham and CLP mice. These results show that the increased inhibition of HCO3 absorption by basolateral LPS in CLP MTALs is mediated through the ERK signaling pathway.

Fig. 7.

Inhibitors of ERK activation eliminate inhibition of HCO3 absorption by basolateral LPS. MTALs from sham (A) and CLP (B) mice were isolated and perfused in vitro 18 h after surgery. Tubules were bathed with 15 μM U0126 or 15 μM PD98059, and then LPS (500 ng/ml) was added to and removed from the bath solution. JHCO3, data points, lines, and P values are as in Fig. 5.

Effect of LPS on ERK phosphorylation in MTALs from sham and CLP mice.

The preceding results suggest that CLP may enhance inhibition of HCO3 absorption by basolateral LPS by upregulating the LPS-induced ERK pathway. Our results show, however, that the baseline p-ERK level is increased in MTALs from CLP mice, raising the question of whether LPS can increase ERK phosphorylation further in CLP MTALs that already have an elevated p-ERK level. MTALs dissected from sham and CLP mice 18 h after surgery were incubated in the absence and presence of LPS for 15 min and then analyzed for ERK phosphorylation by confocal immunofluorescence. The results show that treatment with LPS increased ERK phosphorylation in MTALs from both groups and that the level of phosphorylated ERK was higher in tubules from the CLP mice (Fig. 8A). The intensity of p-ERK staining in the presence of LPS was increased by 30% in MTALs from CLP mice compared with sham controls (Fig. 8B). These results show that LPS stimulates ERK phosphorylation in CLP MTALs despite the presence of an elevated baseline p-ERK level and that the effect of CLP to enhance inhibition of HCO3 absorption by basolateral LPS is associated with an increase in LPS-induced ERK phosphorylation.

Fig. 8.

Effect of LPS on ERK phosphorylation in MTALs from sham and CLP mice. A: MTALs dissected from sham and CLP mice 18 h after surgery were incubated in vitro for 15 min at 37°C in the absence (control) and presence of LPS (500 ng/ml). Tubules were fixed and stained with p-ERK antibody and analyzed by confocal immunofluorescence as in Fig. 3A. LPS increased ERK phosphorylation in MTALs from both sham and CLP mice despite the elevated baseline p-ERK level in the CLP tubules. Images are representative of at least 7 tubules of each type. B: intensity of p-ERK staining in sham and CLP MTALs in the presence of LPS. Fluorescence intensity was quantified as described in materials and methods and expressed in arbitrary fluorescence units. LPS-induced p-ERK level was higher in MTALs from CLP mice. Bars are means ± SE. *P < 0.05 vs. sham.

CLP increases basolateral TLR4 expression in the MTAL.

We have demonstrated previously that basolateral LPS stimulates ERK and inhibits HCO3 absorption in the MTAL through TLR4 (24, 66). We therefore tested whether the enhanced responsiveness of CLP MTALs to basolateral LPS was associated with a change in TLR4 expression. MTALs dissected from sham and CLP mice 18 h after surgery were stained with anti-TLR4 antibody and analyzed by confocal immunofluorescence. As shown in Fig. 9, the intensity of TLR4 staining along the basolateral membrane was increased in MTALs from the CLP mice. Thus the enhanced ability of basolateral LPS to inhibit HCO3 absorption in MTALs from CLP mice is associated with increased expression of TLR4 in the basolateral membrane domain.

Fig. 9.

Basolateral expression of Toll-like receptor 4 (TLR4) is increased in MTALs from CLP mice. A: MTALs dissected from sham and CLP mice 18 h after surgery were stained with anti-TLR4 antibody and analyzed by confocal immunofluorescence as described in Fig. 3A. TLR4 labeling along the basolateral membrane domain was increased in MTALs from CLP mice (arrowheads). Images are representative of at least 8 tubules of each type. B: intensity of TLR4 staining in the basolateral membrane domain was quantified for experiments in A as described in materials and methods and is expressed as a percentage of the sham level measured in the same experiment. Bars are means ± SE. *P < 0.05 vs. sham.

Pretreatment with ERK inhibitor does not prevent CLP-induced enhancement of inhibition by LPS.

Based on our finding that pretreatment with PD98059 prevented the CLP-induced decrease in basal HCO3 absorption rate, we tested whether pretreatment with the inhibitor would influence the effect of CLP to upregulate inhibition by basolateral LPS. Mice were administered PD98059 1 h before sham or CLP surgery. MTALs from PD98059-treated mice were isolated 18 h after surgery, and the effect of basolateral LPS on HCO3 absorption was examined. Adding LPS to the bath decreased HCO3 absorption by 20% (from 15.1 ± 0.8 to 12.1 ± 0.7 pmol·min−1·mm−1; P < 0.001) in MTALs from PD98059-treated sham mice compared with a decrease of 30% (from 14.7 ± 0.4 to 10.3 ± 0.4 pmol·min−1·mm−1; P < 0.001) in MTALs from PD98059-treated CLP mice (Fig. 10A). Similar to results in mice not receiving PD98059, both the fractional and absolute decreases in HCO3 absorption induced by bath LPS were significantly higher in MTALs from PD98059-treated CLP mice than in tubules from PD98059-treated sham controls (Fig. 10B). In addition, the effect of bath LPS to inhibit HCO3 absorption in MTALs from PD98059-treated CLP mice was eliminated in tubules bathed with the MEK inhibitor (Fig. 10C). Thus the effect of CLP to enhance inhibition of HCO3 absorption by basolateral LPS through ERK activation was not prevented by pretreatment of the mice with a MEK/ERK inhibitor. These results show that pretreatment of mice with a MEK/ERK inhibitor does not prevent the ability of CLP to upregulate the basolateral TLR4-ERK pathway through which LPS inhibits HCO3 absorption in the MTAL. In addition, they show that the CLP-induced increase in basal ERK phosphorylation is not necessary for upregulation of the basolateral LPS pathway in CLP MTALs.

Fig. 10.

Effects of basolateral LPS on HCO3 absorption in MTALs from sham and CLP mice pretreated with PD98059. Mice were given PD98059 (16 mg/kg ip) 1 h before sham or CLP surgery. A: MTALs from PD98059-treated mice were perfused in vitro 18 h after surgery. Tubules were studied under basal conditions, and then LPS (500 ng/ml) was added to and removed from the bath solution. JHCO3, data points, lines, and P values are as in Fig. 5. Mean values are given in results. B: fractional and absolute decreases in HCO3 absorption induced by bath LPS in experiments in A. Bars are means ± SE. *P < 0.05 vs. PD98059 + sham. C: MTALs isolated from PD98059-treated CLP mice 18 h after surgery were perfused in vitro with 15 μM PD98059 in the bath, and then LPS (500 ng/ml) was added to and removed from the bath solution. JHCO3, data points, lines, and P value are as in Fig. 5.

CLP does not affect inhibition of HCO3 absorption by aldosterone in the MTAL.

Based on our finding that CLP enhances ERK-dependent inhibition of HCO3 absorption by LPS, further studies were carried out to test whether CLP may upregulate the response of the MTAL to other stimuli that inhibit HCO3 absorption through ERK activation. MTALs from sham and CLP mice were perfused in vitro to examine the effect of aldosterone, which inhibits HCO3 absorption in the MTAL through an ERK-dependent pathway distinct from that activated by LPS (65, 66). Adding aldosterone to the bath decreased HCO3 absorption by 29 ± 2% in MTALs from sham mice and by 31 ± 3% in MTALs from CLP mice (P = NS) (Fig. 11). These results show that the inhibition of HCO3 absorption by aldosterone was unaffected by CLP. Thus the enhanced responsiveness of CLP MTALs to basolateral LPS is a selective effect of sepsis to upregulate the LPS-induced TLR4-ERK pathway and not the result of nonspecific activation of ERK signaling in CLP MTALs.

Fig. 11.

Effect of aldosterone on HCO3 absorption. MTALs from sham and CLP mice 18 h after surgery were perfused in vitro under basal conditions, and then aldosterone (1 nM) was added to and removed from the bath solution. JHCO3, data points, lines, and P value are as in Fig. 5. Mean values are given in results.

DISCUSSION

Kidney dysfunction contributes to sepsis pathogenesis through the loss of fluid, electrolyte, and metabolic homeostasis, and the development of renal insufficiency increases mortality rate in septic patients to very high levels (12, 35, 42, 51, 54, 58). Identifying molecular events through which sepsis impairs renal tubule function would aid in identifying potential therapeutic targets to preserve or restore kidney function in septic patients. One impediment to understanding the pathophysiology of sepsis-induced renal tubule dysfunction is that there have been no direct studies of tubule transport in a clinically relevant sepsis model. In the present study, we used a mouse CLP model in combination with methods for study of isolated renal tubules to directly examine the effects of sepsis on the transport function of the MTAL. Our results show that sepsis can decrease HCO3 absorption in the MTAL through two distinct mechanisms: 1) a decrease in the intrinsic capacity of the tubules to absorb HCO3, and 2) an enhanced ability of bacterial LPS to inhibit HCO3 absorption, mediated by upregulation of the basolateral membrane TLR4 signaling pathway. Thus sepsis can impair renal tubule function both by altering acute responses to inflammatory stimuli and by causing chronic adaptations that impact tubule transport capacity. We demonstrate further that the two inhibitory mechanisms are additive, such that the combination of the two effects reduces the HCO3 absorption rate in MTALs from septic mice to less than half that observed in normal controls. Moreover, both mechanisms depend on ERK activation, suggesting that the ERK pathway may represent a therapeutic target for sepsis-induced renal tubule dysfunction. The effects of sepsis to reduce renal tubule HCO3 absorption are maladaptive since they would impair the ability of the kidneys to correct systemic metabolic acidosis that contributes to sepsis pathogenesis through a variety of mechanisms, including adverse effects on the cardiovascular system and increasing circulating levels of proinflammatory mediators (5, 17, 33, 36, 38, 58). Increased mortality in septic patients was reported recently to correlate with failure to clear metabolic acidosis related to reduced renal function (49).

Absorption of HCO3 by the MTAL undergoes chronic regulation in response to a number of pathophysiological conditions, including changes in acid-base and sodium balance (19, 20, 26, 37). These adaptations contribute to changes in renal net acid excretion that aid in maintaining acid-base homeostasis (20). The results of the present study show that adaptive changes in MTAL HCO3 absorption also can contribute to disease pathogenesis. Sepsis decreased the basal HCO3 absorption rate despite the presence of metabolic acidosis that, by itself, induces an increase in MTAL HCO3 absorptive capacity (19, 20, 37). The effect of sepsis reflects an adaptive change in the intrinsic transport capacity of the tubule cells because the decreased HCO3 absorption rate persists when the tubules are removed from the septic environment and studied under standard conditions in vitro. In contrast to HCO3 absorption, sepsis had no effect on transepithelial voltage, a correlate of NaCl absorption rate. Thus the adaptation appears to involve a selective effect on HCO3 absorptive capacity and not a nonspecific metabolic or toxic effect of sepsis on the MTAL cells. This conclusion is supported further by the findings that MTALs from CLP mice exhibit normal or enhanced responses to two different stimuli (aldosterone and LPS) and that inhibition of transport by these factors is fully reversible. Thus our functional data in the MTAL do not provide evidence to support a sepsis-induced defect in mitochondrial function or oxidative metabolism, factors recently suggested to contribute to renal tubule dysfunction in an LPS injection endotoxemia model (61). Our finding that CLP has no effect on transepithelial voltage suggests that baseline NaCl absorption is preserved in MTALs from septic mice under the conditions of our experiments. Previous studies have shown, however, that TNF-α and IL-1 inhibit ouabain-sensitive 86Rb uptake by MTAL cells and that TNF-α can inhibit apical Na+-K+-2Cl cotransport (NKCC2) through a PGE2-dependent mechanism (15, 16). These findings suggest that MTAL NaCl absorption may be reduced during sepsis in vivo through the action of locally produced or circulating cytokines. Alternatively, TNF-α has been shown to decrease NOS3 expression in MTAL cells, which would promote increased NaCl absorption by reducing nitric oxide levels (52). Further direct studies of the effects of sepsis, inflammatory mediators, and TLR4 agonists on MTAL NaCl absorption are needed.

The sepsis-induced decrease in HCO3 absorptive capacity is associated with an increase in ERK phosphorylation in the MTAL. The latter finding is consistent with results of previous studies showing that CLP increases ERK phosphorylation in lung and liver (72). The increase in ERK phosphorylation occurs in the absence of a change in total ERK expression, suggesting that sepsis activates ERK by altering the relative rates of phosphorylation/dephosphorylation events. The systemic factors and signaling pathways responsible for sepsis-induced ERK activation remain to be determined. However, pretreating CLP mice with a MEK inhibitor to suppress ERK activation prevents the sepsis-induced decrease in HCO3 absorptive capacity, such that the basal rate of HCO3 absorption in MTALs from septic mice is restored to a value not different from sham controls. These findings do not establish unequivocally that increased ERK activity in the MTAL mediates the decrease in HCO3 absorptive capacity. For example, we cannot rule out the possibility that the MEK inhibitor acts on other cell types to reduce ERK-dependent production of an inflammatory mediator(s) that secondarily decreases MTAL HCO3 absorptive capacity. Nevertheless, our finding of a direct correlation between increased ERK phosphorylation and decreased HCO3 absorption rate in MTALs from septic animals, combined with our previous studies demonstrating directly that activation of ERK decreases MTAL HCO3 absorption (23, 65, 68), strongly suggests that the two effects are causally related. This view is supported further by studies of the transport mechanism underlying the decrease in HCO3 absorption in CLP tubules. Basolateral NHE1 is an important determinant of HCO3 absorption rate in the MTAL (21, 23, 26, 29, 63), and a major mechanism by which ERK inhibits HCO3 absorption is through inhibition of NHE1 (21, 29, 63, 64, 68). Our current data support a role for NHE1 in mediating the reduced HCO3 absorptive capacity in MTALs from septic mice. Bath amiloride, which inhibits HCO3 absorption in the MTAL through inhibition of basolateral NHE1 (21, 23, 29, 63, 64), decreased HCO3 absorption in MTALs from sham mice but had no effect in MTALs from CLP mice. The most likely explanation for this result is that NHE1 activity is reduced or absent in MTALs from the septic mice, which decreases the basal HCO3 absorption rate and precludes inhibition of HCO3 absorption by bath amiloride because NHE1 activity is already inhibited. Taken together, these findings support a mechanism whereby sepsis increases ERK activity in the MTAL, which reduces basal HCO3 absorptive capacity through inhibition of basolateral NHE1. Further direct studies of the effects of CLP on basolateral Na+/H+ exchange activity and NHE1 expression in the MTAL will be required to test this hypothesis. An additional possibility is that sepsis may act independently of NHE1 to reduce the activity of NHE3, the apical Na+/H+ exchanger that mediates H+ secretion necessary for MTAL HCO3 absorption (1, 7, 20, 28, 37, 67). Studies using the LPS injection model of endotoxemia found that expression of NHE3 was decreased after 6 to 12 h in whole kidneys of mice (62) and in the inner stripe of the outer medulla of kidneys from rats (50). Whether expression of renal transport proteins is altered in more clinically relevant models of sepsis (11), and whether changes in expression of NHE3 may contribute to the decreased HCO3 absorptive capacity of the MTAL in CLP, will be important questions for future investigation.

In addition to reducing the basal HCO3 absorption rate, sepsis enhanced the ability of basolateral LPS to inhibit HCO3 absorption. We have shown previously that basolateral LPS inhibits HCO3 absorption in the MTAL through TLR4-mediated activation of ERK (24, 66). The results of the present study show that this innate immune pathway is upregulated by sepsis. Sepsis increased expression of TLR4 in the basolateral membrane of the mouse MTAL. This finding is consistent with the results of previous studies showing that CLP increased TLR4 expression in several nephron segments in rats, including the proximal tubule and thick ascending limb, where it may mediate the production of proinflammatory mediators (13, 14). In the MTAL, the increased expression of TLR4 would contribute to sepsis pathogenesis by inducing intracellular signals that impair tubule transport. The level of ERK phosphorylation in the presence of LPS is higher in MTALs from septic animals, an effect presumably facilitated by increased TLR4 expression, and the increased ability of LPS to inhibit HCO3 absorption in CLP tubules depends on ERK activation. Two additional aspects of CLP-induced upregulation of TLR4-ERK signaling are noteworthy. First, CLP had no effect on ERK-dependent inhibition of HCO3 absorption by aldosterone. This shows that upregulation of ERK signaling in the MTAL is selective for the LPS-induced TLR4 pathway and is not the result of a nonspecific effect of sepsis to enhance ERK signaling or the responsiveness of transport proteins involved in HCO3 absorption to ERK activation. Second, LPS is able to activate ERK in CLP MTALs despite the presence of an already elevated basal phospho-ERK level. This shows that the ability of LPS to stimulate ERK and inhibit HCO3 absorption is not limited in the presence of other stimuli that activate the ERK pathway; that is, the stimulation of ERK by LPS through TLR4 is additive to increases in ERK activity induced by other sepsis-associated factors. These findings have important implications for disease pathogenesis since they reveal the presence of specialized mechanisms for upregulation of cell signals during sepsis, whereby bacterial molecules can activate cell signaling pathways to impair tubule function despite the presence of other physiological or pathophysiological stimuli that converge on and activate the same signaling pathways. The mechanisms responsible for this ERK signal specificity in the MTAL remain to be determined but may include coupling of the TLR4 receptor complex to distinct ERK pools, activation of ERK through different upstream mediators, or mechanisms for graded ERK responses in which different inputs activate ERK with different thresholds (2, 4, 48, 66).

A key finding of our study is that the effects of sepsis to decrease basal HCO3 absorption rate and to increase inhibition of HCO3 absorption by LPS are additive. Moreover, the two mechanisms can be induced independently, since upregulation of the basolateral TLR4 pathway is maintained under conditions in which the adaptive decrease in HCO3 absorption is prevented by pretreatment with a MEK inhibitor. We suggest that the additivity of the two sepsis-induced mechanisms is likely a result of specificity of ERK signaling in the MTAL whereby the ERK pathway can be targeted to inhibit different Na+/H+ exchangers in response to different stimuli. We have shown previously that activation of ERK can inhibit HCO3 absorption in the MTAL through two distinct mechanisms: 1) direct coupling of the ERK pathway to inhibition of basolateral NHE1, which results secondarily in inhibition of apical NHE3 (21, 29, 63, 64, 68); and 2) direct coupling of the ERK pathway to inhibition of NHE3, independent of NHE1 (22, 65). Our findings suggest that these two mechanisms may underlie the additive effects of sepsis on HCO3 absorption. The inhibition of HCO3 absorption by basolateral LPS is mediated through direct coupling of the ERK pathway to inhibition of NHE3 (66). Conversely, as discussed above, the sepsis-induced decrease in basal HCO3 absorption rate appears to involve inhibition of NHE1. The effects of ERK to inhibit HCO3 absorption through direct coupling to NHE1 and NHE3 are additive, as demonstrated previously (22). These proposed mechanisms can explain our findings that the effects of sepsis to decrease basal HCO3 absorption and to enhance inhibition by LPS are additive and independent and that both effects can be prevented by targeting the ERK pathway. Important goals for future work will be to understand how the ERK pathway is specifically targeted to regulate different Na+/H+ exchangers in the same cell and to examine directly the effects of CLP on ERK-dependent regulation of NHE1 and NHE3 in the MTAL.

In previous studies, we uncovered novel features of LPS signaling in the MTAL that may impact how sepsis alters renal tubule function. First, we identified a sidedness to LPS receptor signaling, whereby LPS inhibits HCO3 absorption in the MTAL through the activation of different TLR4-dependent signaling pathways in the basolateral and apical membranes (24). These findings raise the possibility that the different TLR4 signaling pathways in the two membranes may subserve different functions in the renal tubule innate immune response and could be differentially regulated by sepsis or other inflammatory conditions. Second, we identified a novel dependence of basolateral TLR4 signaling on TLR2, in which the effects of basolateral LPS to activate ERK and inhibit HCO3 absorption in the MTAL involve an interaction between TLR4 and TLR2 in the basolateral membrane (27). These findings raise the possibility that upregulation of the LPS-induced ERK pathway in the MTAL by sepsis could involve increased expression of basolateral TLR2 in addition to TLR4. Alternatively, sepsis could alter the molecular association between TLR4 and TLR2 in a way that enhances LPS-induced ERK activation through the two receptors. These findings suggest new molecular mechanisms through which sepsis may impact renal tubule signaling pathways regulating ion transport and inflammation.

Our finding that TLR4 signaling linked to inhibition of HCO3 absorption is upregulated in MTALs from CLP mice has implications for inflammation-associated kidney dysfunction that extend beyond LPS and sepsis. In addition to its role as the LPS receptor, TLR4 is involved in mediating signaling responses induced by a variety of endogenous molecules, including heat shock proteins, the nuclear-binding protein high-mobility group box 1 protein (HMGB1), and extracellular matrix components such as hyaluronan and biglycan (6, 32, 45). These endogenous “danger” molecules are released by stressed or damaged cells, and activation of TLR4 on renal tubule cells by these ligands is thought to play a role in mediating inflammatory kidney injury in sepsis as well as a number of noninfectious conditions, including ischemia-reperfusion, glomerular diseases, nephrotoxic injury, renal transplant rejection, and obstruction (8, 31, 34, 39, 41, 43, 44, 59, 70, 71). Our findings raise the possibility that endogenous molecules may activate TLR4-dependent cell signals that impair renal tubule transport and that these signals may be amplified by sepsis through upregulation of the basolateral TLR4 pathway. In addition, TLR4 expression is increased in renal tubule segments by ischemia-reperfusion, diabetes, and nephrotoxic drugs (34, 43, 44, 6971). Thus it is possible that upregulation of the TLR4-ERK pathway we identified in sepsis may enable endogenous ligands to induce renal tubule transport defects in these noninfectious inflammatory conditions. Our findings raise several important questions for future study, including whether endogenous “danger” molecules can activate TLR4-mediated signals that impair ion transport in the MTAL and how endogenous ligands may interact with bacterial ligands to impact TLR4 signaling and renal tubule function during infection.

In summary, we have used a CLP model to identify immunopathogenic mechanisms by which sepsis impairs the function of renal tubules. The results show that sepsis reduces the ability of the MTAL to absorb HCO3 through two distinct mechanisms. First, sepsis induces an adaptive decrease in the intrinsic capacity of the MTAL to absorb HCO3. This effect appears to involve loss of function of basolateral membrane Na+/H+ exchange. Second, sepsis enhances the ability of LPS to inhibit HCO3 absorption. This effect is mediated through upregulation of TLR4 signaling in the basolateral membrane. The two inhibitory mechanisms are additive and thus function in a two-hit capacity to impair MTAL transport. In addition, both effects of sepsis to reduce basal HCO3 absorptive capacity and to upregulate TLR4-mediated inhibition by LPS depend on ERK and are eliminated by interventions that prevent ERK activation. Thus selective targeting of the ERK pathway represents a potential therapeutic approach for treatment or prevention of sepsis-induced renal tubule dysfunction.

GRANTS

This work was supported by National Institute of Diabetes and Digestive and Kidney Diseases Grant DK-038217.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the author(s).

AUTHOR CONTRIBUTIONS

Author contributions: B.A.W., E.R.S., and D.W.G. conception and design of research; B.A.W., T.G., and E.R.S. performed experiments; B.A.W., E.R.S., and D.W.G. analyzed data; B.A.W., E.R.S., and D.W.G. interpreted results of experiments; B.A.W. and D.W.G. prepared figures; B.A.W., E.R.S., and D.W.G. approved final version of manuscript; D.W.G. drafted manuscript.

ACKNOWLEDGMENTS

We thank Daniela Herzig for assistance in experiments measuring kidney cytokines, and we thank Geping Fang for technical support.

Present address of E. R. Sherwood: Dept. of Anesthesiology, Vanderbilt Univ. Medical Center, Nashville, TN.

REFERENCES

  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
  24. 24.
  25. 25.
  26. 26.
  27. 27.
  28. 28.
  29. 29.
  30. 30.
  31. 31.
  32. 32.
  33. 33.
  34. 34.
  35. 35.
  36. 36.
  37. 37.
  38. 38.
  39. 39.
  40. 40.
  41. 41.
  42. 42.
  43. 43.
  44. 44.
  45. 45.
  46. 46.
  47. 47.
  48. 48.
  49. 49.
  50. 50.
  51. 51.
  52. 52.
  53. 53.
  54. 54.
  55. 55.
  56. 56.
  57. 57.
  58. 58.
  59. 59.
  60. 60.
  61. 61.
  62. 62.
  63. 63.
  64. 64.
  65. 65.
  66. 66.
  67. 67.
  68. 68.
  69. 69.
  70. 70.
  71. 71.
  72. 72.
View Abstract