IκB Kinase Inhibitor Attenuates Sepsis-Induced Cardiac Dysfunction in CKD

Abstract

Patients with CKD requiring dialysis have a higher risk of sepsis and a 100-fold higher mortality rate than the general population with sepsis. The severity of cardiac dysfunction predicts mortality in patients with sepsis. Here, we investigated the effect of preexisting CKD on cardiac function in mice with sepsis and whether inhibition of IκB kinase (IKK) reduces the cardiac dysfunction in CKD sepsis. Male C57BL/6 mice underwent 5/6 nephrectomy, and 8 weeks later, they were subjected to LPS (2 mg/kg) or sepsis by cecal ligation and puncture (CLP). Compared with sham operation, nephrectomy resulted in significant increases in urea and creatinine levels, a small (P<0.05) reduction in ejection fraction (echocardiography), and increases in the cardiac levels of phosphorylated IκBα, Akt, and extracellular signal–regulated kinase 1/2; nuclear translocation of the NF-κB subunit p65; and inducible nitric oxide synthase (iNOS) expression. When subjected to LPS or CLP, compared with sham-operated controls, CKD mice exhibited exacerbation of cardiac dysfunction and lung inflammation, greater increases in levels of plasma cytokines (TNF-α, IL-1β, IL-6, and IL-10), and greater increases in the cardiac levels of phosphorylated IKKα/β and IκBα, nuclear translocation of p65, and iNOS expression. Treatment of CKD mice with an IKK inhibitor (IKK 16; 1 mg/kg) 1 hour after CLP or LPS administration attenuated these effects. Thus, preexisting CKD aggravates the cardiac dysfunction caused by sepsis or endotoxemia in mice; this effect may be caused by increased cardiac NF-κB activation and iNOS expression.

Sepsis is a systemic dysregulated inflammatory response to an infection, which when excessive, may progress to multiple organ failure and death.¹ More than 40% of cases of sepsis have cardiovascular impairment,² and the overall mortality in patients with sepsis who have myocardial dysfunction rises from 40% to 70%.³ The lack of translatability of preclinical findings to patients with sepsis has many possible reasons, including interventions given relatively late and a great degree of heterogeneity in the patient population, which often has comorbidities, including diabetes, CKD, or both.^4–7 CKD is a growing public health burden, with an increasing number of patients receiving maintenance dialysis.⁸ Cardiovascular disease is the leading cause of death in patients with CKD.⁹ The cardiac injury caused by ischemia-reperfusion is greater in uremic rats compared with nonuremic controls.¹⁰ Patients with CKD requiring dialysis have a higher risk of infection and sepsis¹¹ because of uremia–induced immune deficiency,^12–14 significant comorbidities, and the dialysis procedure itself.¹⁵ After infected, patients on dialysis with sepsis have an approximately 100-fold higher mortality rate compared with the general population with sepsis.¹⁶ It is possible that alterations in cardiac function (at baseline, in response to sepsis, or both) play a crucial role in the increased risk of death in patients with CKD and sepsis.

Upregulation of NF-κB has been linked to the development of cardiac dysfunction after the onset of sepsis.^17,18 Physiologically, inhibitor of κBα (IκBα) inactivates NF-κB by sequestering NF-κB as an inactive complex in the cytoplasm.^19,20 Phosphorylation of IκBα by inhibitor of κB kinase (IKK) dissociates IκBα from NF-κB, which liberates NF-κB to enter the nucleus and activates the expression of NF-κB target genes.²⁰ Inhibition of IKK²¹ attenuates sepsis–induced multiple organ dysfunction/injury in mice.²² It is, however, unknown whether preexisting CKD augments the cardiac dysfunction in sepsis and whether excessive activation of NF-κB drives cardiac dysfunction in animals with CKD and sepsis.

Results

Characterization of Organ Dysfunctions and Blood Tests in Mice That Underwent Subtotal (5/6th) Nephrectomy

Compared with a sham procedure, subtotal (5/6th) nephrectomy (SNX) resulted in significantly higher plasma urea and creatinine concentrations; this was paralleled by a mild cardiomyopathy indicated by slight but significant reductions in percentage of ejection fraction (EF), fractional shortening (FS), and fractional area change (FAC). CKD mice exhibited a significantly higher mean arterial BP (MABP), greater heart weight, and greater heart weight-to-body weight ratio (a surrogate marker for myocardial hypertrophy²³; P<0.05) (Supplemental Table 1). Additionally, there was an increase in interventricular septum thickness in CKD mice (P<0.05), but no difference was observed in left ventricular dimensions (left ventricular internal diastolic dimension or left ventricular end diastolic volume; P>0.05) (Supplemental Table 1), indicating the development of concentric hypertrophy of CKD hearts.

Additionally, full blood analysis indicated the development of anemia and an increase in the neutrophil-to-lymphocyte ratio in CKDs (P<0.05) (Supplemental Table 1). Most notably, CKD mice had elevated plasma levels of the (mainly proinflammatory) cytokines IL-1β and keratinocyte-derived cytokine (Supplemental Figure 1, Supplemental Table 1), TNF-α, IL-6, and IL-10 (Supplemental Figure 1), indicating that CKD caused mild systemic inflammation.

Preexisting CKD Augmented the Cardiac Dysfunction Caused by Low–Dose LPS Administration

In CKD sham animals, low-dose LPS (2 mg/kg) had no effect on percentage of EF, FAC, and FS (P>0.05) (Figure 1, A–D, Supplemental Figure 2); however, in CKD mice, low-dose LPS induced significant reductions in percentage of EF, FAC, and FS (P<0.05) (Figure 1, A–D, Supplemental Figure 2), indicating the development of a clear and significant cardiac dysfunction in vivo.

Figure 1.

Preexisting CKD augments the cardiac dysfunction induced by low–dose LPS administration or CLP. (A–D) CKD sham or CKD mice received either LPS (2 mg/kg) or PBS (5 ml/kg) intraperitoneally. Cardiac function was assessed at 18 hours. (A) Representative M–mode echocardiograms and percentages of (B) EF, (C) FAC, and (D) FS. The following groups were studied: CKD sham and PBS (n=6), CKD and PBS (n=7), CKD sham and LPS (2 mg/kg; n=7), and CKD and LPS (2 mg/kg; n=7). (E–H) CKD sham or CKD mice were subjected to CLP or sham-operated surgery. Cardiac function was assessed at 24 hours. (E) Representative M–mode echocardiograms and percentages of (F) EF, (G) FAC, and (H) FS. The following groups were studied: CKD sham and sham operated (n=6), CKD and sham operated (n=7), CKD sham and CLP (n=7), and CKD and CLP (n=7). All data are represented as means±SEMs. Data were analyzed by one-way ANOVA followed by Bonferroni post hoc test. ^★P<0.05 versus the CKD sham group with respective treatment; ^#P<0.05 versus the respective PBS or sham-operated group.

Preexisting CKD Augmented the Cardiac Dysfunction Caused by Cecal Ligation and Puncture

The murine model of cecal ligation and puncture (CLP) with fluid resuscitation and antibiotics treatment offers a clinically relevant model of abdominal polymicrobial human sepsis. CLP–induced cardiac dysfunction was only observed in 8-month-old mice but not in young mice.¹⁸ As previously reported,¹⁸ CLP had no significant effect on cardiac parameters in young mice (P>0.05) (Figure 1, E–H, Supplemental Figure 3). However, in CKD mice, CLP caused significant reductions in percentage of EF, FAC, and FS (P<0.05) (Figure 1, E–H, Supplemental Figure 3), indicating the development of a pronounced cardiac dysfunction in vivo. The degree of systolic dysfunction in young CKD mice with CLP was similar to the cardiac dysfunction reported previously in old (8 months) mice with CLP.¹⁸ The cardiac dysfunction in CKD/CLP mice was paralleled with a reduced physical activity (P<0.05) (Figure 2A). The drop in MABP was slightly greater in CKD/CLP mice compared with sham mice (P<0.05) (Figure 2B); however, the drastic cardiac dysfunction observed in CKD/CLP animals cannot be attributed to such a small change in BP, indicating that the cardiac dysfunction might not be primarily dependent on MABP.

Figure 2.

CLP and/or IKK 16 treatment causes a significant change in activity (Δactivity), but a slight change in MABP (ΔMABP) in CKD mice. Radiotelemetric recording of (A) Δactivity and (B) conscious ΔMABP of CKD sham (black) or CKD (white and red) mice subjected to CLP. After 1 hour (at 3:00 p.m.) of CLP, CKD mice were injected with vehicle (black) or IKK 16 (red; n=3–4 per group). All data are represented as means±SEMs. (A) Data were analyzed by one-way ANOVA followed by Bonferroni post hoc test. (B) Data were analyzed by two-way ANOVA followed by Bonferroni post hoc test. ^★P<0.05 versus the CKD, CLP, and vehicle group.

Increases in the Phosphorylation of IKKα/β, the Phosphorylation of IκBα, the Nuclear Translocation of p65 NF-κB, and the Inducible Nitric Oxide Synthase Expression in Hearts of Mice with CKD Subjected to Low–Dose LPS Administration or CLP

To gain a better mechanical insight into the augmented sepsis–associated cardiac dysfunction in CKD mice, we investigated the effects of preexisting CKD on signaling events in mouse hearts subjected to LPS or CLP. Compared with PBS–treated or sham–operated CKD sham mice, PBS–treated or sham–operated CKD mice exhibited significantly higher degrees of cardiac phosphorylation of IKKα/β on Ser^176/180, subsequent phosphorylation of IκBα on Ser^32/36, nuclear translocation of p65 NF-κB, and inducible nitric oxide synthase (iNOS) expression (P<0.05) (Figures 3, A–D and 4, A–D). Exposure of CKD sham mice to low-dose LPS or CLP had no significant effect on any of the above signaling pathways (P>0.05) (Figures 3, A–D and 4, A–D). However, LPS or CLP further increased cardiac phosphorylation of IKKα/β and IκBα, nuclear translocation of p65, and iNOS expression (P<0.05) (Figures 3, A–D and 4, A–D) to profound degrees in CKD mice.

Figure 3.

Low–dose LPS administration increases the phosphorylation of IKKα/β and IκBα, the nuclear translocation of p65 NF-κB and the iNOS expression, but causes no change in the phosphorylation of Akt or ERK1/2 in hearts of mice with CKD. CKD sham or CKD mice received either LPS (2 mg/kg) or PBS (5 ml/kg) intraperitoneally. Signaling events in heart tissue were assessed at 18 hours. Densitometric analysis of the bands is expressed as relative OD of (A) phosphorylated IKKα/β (pSer^176/180) corrected for the corresponding total IKKα/β content and normalized using the related sham band, (B) phosphorylated IκBα (pSer^32/36) corrected for the corresponding total IκBα content and normalized using the related sham band, (C) NF-κB p65 subunit levels in both cytosolic and nuclear fractions expressed as a nucleus-to-cytosol ratio normalized using the related sham bands, (D) iNOS expression corrected for the corresponding tubulin band, (E) phosphorylated Akt (pSer⁴⁷³) corrected for the corresponding total Akt content and normalized using the related sham band, and (F) ERK1/2 phosphorylation corrected for the corresponding total ERK1/2 content and normalized using the related sham band. Each analysis is from a single experiment and representative of three separate experiments. Data are expressed as means±SEMs for n number of observations. Data were analyzed by one-way ANOVA followed by Bonferroni post hoc test. ph, Phospho. ^★P<0.05 versus the CKD sham group with respective treatment; ^#P<0.05 versus the respective PBS group.

Figure 4.

CLP increases the phosphorylation of IKKα/β and IκBα, the nuclear translocation of p65 NF-κB and the iNOS expression, but causes no change in the phosphorylation of Akt or ERK1/2 in hearts of mice with CKD. CKD sham or CKD mice were subjected to CLP or sham-operated surgery. Signaling events in heart tissue were assessed at 24 hours. Densitometric analysis of the bands is expressed as relative OD of (A) phosphorylated IKKα/β (pSer^176/180) corrected for the corresponding total IKKα/β content and normalized using the related sham band, (B) phosphorylated IκBα (pSer^32/36) corrected for the corresponding total IκBα content and normalized using the related sham band, (C) NF-κB p65 subunit levels in both cytosolic and nuclear fractions expressed as a nucleus-to-cytosol ratio normalized using the related sham bands, (D) iNOS expression corrected for the corresponding tubulin band, (E) phosphorylated Akt (pSer⁴⁷³) corrected for the corresponding total Akt content and normalized using the related sham band, and (F) ERK1/2 phosphorylation corrected for the corresponding total ERK1/2 content and normalized using the related sham band. Each analysis is from a single experiment and representative of three separate experiments. Data are expressed as means±SEMs for n number of observations. Data were analyzed by one-way ANOVA followed by Bonferroni post hoc test. ph, Phospho. ^★P<0.05 versus the CKD sham group with respective treatment; ^#P<0.05 versus the respective sham–operated group.

Effects of Low–Dose LPS Administration or CLP on the Phosphorylation of Akt and Extracellular Signal–Regulated Kinase 1/2 in Hearts of Mice with CKD

Compared with PBS–treated or sham–operated CKD sham mice, PBS–treated or sham–operated CKD mice showed significantly higher degrees of cardiac phosphorylation of Akt on Ser⁴⁷³ and extracellular signal–regulated kinase 1/2 (ERK1/2) on Tyr²⁰² and Tyr²⁰⁴, respectively (P<0.05) (Figures 3, E and F and 4, E and F). CKD sham or CKD mice subjected to LPS or CLP showed no significant change in the degree of phosphorylation of Akt or ERK1/2 (P>0.05) (Figures 3, E and F and 4, E and F).

Preexisting CKD Increases Severity of Renal Dysfunction and Hepatocellular Injury Caused by Low–Dose LPS Administration or CLP

In CKD sham animals, septic insults induced by either low-dose LPS or CLP had no significant effect on plasma urea, creatinine, or alanine aminotransferase (ALT) level (P>0.05) (Table 1); however, in CKD mice, low-dose LPS further increased plasma urea, creatinine, and ALT levels to profound degrees (P<0.05) (Table 1). CLP resulted in significant increases in plasma urea and ALT levels (P<0.05) (Table 1), indicating the augmentation of renal dysfunction and hepatocellular injury, respectively.

Table 1.

Effects of low–dose LPS (2 mg/kg) administration or polymicrobial sepsis induced by CLP on renal dysfunction and hepatocellular injury in mice with CKD

Parameter	CKD Sham		CKD
	PBS	LPS, 2 mg/kg	PBS	LPS, 2 mg/kg
No.	6	7	7	7
Urea, mmol/L	8.26±0.47	16.13±3.88	17.24±1.09a	38.56±2.11a,b
Creatinine, μmol/L	30.22±0.55	30.23±2.35	45.47±2.42a	58.43±2.55a,b
ALT, U/L	27.23±3.01	52.06±2.11	32.16±3.34	83.35±14.11a,b

	Sham Operated	CLP	Sham Operated	CLP
No.	6	6	7	7
Urea, mmol/L	8.08±0.72	13.08±087	17.61±0.66	37.60±6.91a,b
Creatinine, μmol/L	29.22±0.50	27.30±0.93	46.44±2.75	67.43±12.92a
ALT, U/L	23.62±2.90	103.52±15.31	42.44±8.10	287.10±49.86a,b

Plasma urea, creatinine, and ALT levels were assessed at 18 hours in mice subjected to LPS administration and 24 hours in mice that underwent CLP. All data are represented as means±SEMs. Data were analyzed by one-way ANOVA followed by Bonferroni post hoc test.

^aP<0.05 versus the CKD sham group with respective treatment.

^bP<0.05 versus the respective PBS or sham-operated group.

Preexisting CKD Increased Lung Inflammation and Systemic Inflammatory Response Caused by CLP

In CKD sham animals, CLP had no significant effect on lung myeloperoxidase (MPO) activity or plasma inflammatory cytokine levels (TNF-α, IL-1β, IL-6, IL-10, or keratinocyte-derived cytokine; P>0.05) (Figure 5); however, in CKD mice, CLP resulted in significant increases in lung MPO activity and inflammatory cytokine levels (P<0.05) (Figure 5, A–E), indicating an increased neutrophil infiltration in the lung and an enhanced systemic inflammatory response, respectively. No alteration was detected in peritoneal bacteria content between CKD and CKD sham mice after CLP (P>0.05) (Supplemental Figure 4).

Figure 5.

Preexisting CKD increases lung inflammation and systemic inflammatory response induced by CLP. Markers of lung inflammation and systemic response were assessed at 24 hours in mice that underwent CLP. (A) MPO activity in lung tissue, (B) plasma TNF-α concentration, (C) plasma IL-1β concentration, (D) plasma IL-6 concentration, (E) plasma IL-10 concentration, and (F) plasma keratinocyte–derived cytokine (KC) concentration. (A) n=3 per group. (B–F) n=3 for CKD sham and sham-operated groups, and n=5–6 for other groups. All data are represented as means±SEMs. Data were analyzed by one-way ANOVA followed by Bonferroni post hoc test for multiple comparisons or t test for comparisons between two groups. ND, not detected. ^★P<0.05 versus the CKD sham group with respective treatment; ^#P<0.05 versus the respective sham–operated group; ⁺P<0.05 versus the CKD sham group with sham operation.

Inhibition of IKK Attenuated CLP or LPS–Induced Cardiac Dysfunction in Mice with CKD

Compared with sham–operated CKD mice, CKD mice that underwent CLP with vehicle treatment developed significant cardiac dysfunction (P<0.05) (Figure 6, Supplemental Figure 3); this was significantly attenuated by delayed administration of IKK 16 1 hour after CLP (P<0.05) (Figure 6, Supplemental Figure 3). CKD/CLP mice that received IKK 16 were significantly more active than CKD/CLP mice that received vehicle (P<0.05) (Figure 2A). IKK 16 increased MABP in CKD/CLP mice (P<0.05) (Figure 2B). However, IKK 16 did not affect MABP in anesthetized CKD mice (baseline: 84.26±2.08 mmHg versus IKK 16 administration: 82.52±3.83 mmHg; n=3; P>0.05). Therefore, the higher MABP in IKK 16–treated CKD/CLP mice might be caused by improved cardiac function or increased activity (secondary to an overall better health and cardiac performance). No significant change in plasma urea, creatinine, or ALT level was seen with IKK 16 administration (P>0.05) (Supplemental Table 2). Similar protective effects of IKK 16 against cardiac dysfunction were found in CKD mice subjected to LPS administration (Supplemental Figures 2 and 5).

Figure 6.

Inhibition of IKK attenuated CLP-induced cardiac dysfunction in mice with CKD. CKD mice underwent sham-operated surgery or CLP. One hour after CLP, mice were treated with either IKK 16 (1 mg/kg intravenously) or vehicle (2% DMSO). Cardiac function was assessed at 24 hours. (A) Representative M–mode echocardiograms and percentages of (B) EF, (C) FAC, and (D) FS. The following groups were studied: CKD and sham operated (n=7); CKD, CLP, and vehicle (n=7); and CKD, CLP, and IKK 16 (n=7). All data are represented as means±SEMs. Data were analyzed by one-way ANOVA followed by Bonferroni post hoc test. ^★P<0.05 versus the CKD, CLP, and vehicle group.

Effects of IKK Inhibitor on Signaling Events Induced by CLP or LPS in Hearts of CKD Mice

Compared with CKD/CLP mice with vehicle treatment, delayed administration of IKK 16 significantly attenuated the increases in cardiac phosphorylation of IKKα/β and IκBα, nuclear translocation of p65, and iNOS expression (P<0.05) (Figure 7, A–D). Moreover, IKK 16 treatment significantly reduced cardiac phosphorylation of Akt and ERK1/2 (P<0.05) (Figure 7, E and F) in CKD/CLP mice. Similar signaling events were observed in CKD/LPS mice with delayed IKK 16 treatment (Supplemental Figure 6).

Figure 7.

Inhibition of IKK attenuates the increases in cardiac phosphorylation of IKKα/β, IκBα, Akt and ERK1/2, nuclear translocation of p65 and iNOS expression in CKD CLP mice. CKD sham underwent sham-operated surgery, and CKD mice were subjected to CLP or sham-operated surgery. One hour after CLP, CKD mice were treated with either IKK 16 (1 mg/kg intravenously) or vehicle (2% DMSO). Signaling events in heart tissue were assessed at 24 hours. Densitometric analysis of the bands is expressed as relative OD of (A) phosphorylated IKKα/β (pSer^176/180) corrected for the corresponding total IKKα/β content and normalized using the related sham band, (B) phosphorylated IκBα (pSer^32/36) corrected for the corresponding total IκBα content and normalized using the related sham band, (C) NF-κB p65 subunit levels in both cytosolic and nuclear fractions expressed as a nucleus-to-cytosol ratio normalized using the related sham bands, (D) iNOS expression corrected for the corresponding tubulin band, (E) phosphorylated Akt (pSer⁴⁷³) corrected for the corresponding total Akt content and normalized using the related sham band, and (F) ERK1/2 phosphorylation corrected for the corresponding total ERK1/2 content and normalized using the related sham band. Each analysis is from a single experiment and representative of three separate experiments. Data are expressed as means±SEMs for n number of observations. Data were analyzed by one-way ANOVA followed by Bonferroni post hoc test. ph, Phospho. ^★P<0.05 lversus the CKD, CLP, and vehicle group.

Inhibition of IKK Attenuated Lung Inflammation and Systemic Inflammatory Response Caused by CLP or LPS Administration

Treatment of CKD/CLP mice with IKK 16 1 hour after CLP significantly reduced the increases in lung MPO activity and plasma inflammatory cytokine levels (P<0.05) (Figure 8, A–E). Similar protective effects of IKK 16 against lung inflammation and systemic inflammatory response were found in CKD/LPS mice (Supplemental Figure 1). However, IKK 16 treatment had no effect on peritoneal bacteria content in CKD mice after CLP (P>0.05) (Supplemental Figure 4).

Figure 8.

Inhibition of IKK attenuates lung inflammation and systemic inflammatory response induced by CLP in CKD mice. CKD mice underwent CLP or sham-operated surgery. One hour after CLP, CKD mice were treated with either IKK 16 (1 mg/kg intravenously) or vehicle (2% DMSO). Markers of lung inflammation and systemic response were assessed at 24 hours. (A) MPO activity in lung tissue, (B) plasma TNF-α concentration, (C) plasma IL-1β concentration, (D) plasma IL-6 concentration, (E) plasma IL-10 concentration, and (F) plasma keratinocyte–derived cytokine (KC) concentration. (A) n=3 per group, and (B–F) n=5–6 per group. All data are represented as means±SEMs. Data were analyzed by one-way ANOVA followed by Bonferroni post hoc test. ND, not detected. ^★P<0.05 versus the CKD, CLP, and vehicle group.

Discussion

The presence of cardiac dysfunction in patients with sepsis has been linked to a significantly raised mortality rate.³ Patients with CKD also have a significantly higher risk of death after sepsis^15,24; however, the reasons for this higher risk are unclear. This study was designed to elucidate whether preexisting CKD worsens cardiac performance in mice with sepsis and identify (some of) the molecular mechanisms responsible to target/test new therapeutic interventions to reduce cardiac dysfunction in mice with CKD and sepsis.

In mice with SNX for 8 weeks (without sepsis), we found a small but significant impairment in systolic function (EF) and left ventricular hypertrophy (LVH), indicating the development of a cardiorenal syndrome (type 4 as defined by the Acute Dialysis Quality Initiative team).²⁵ This result is consistent with a previous study revealing the presence of impaired cardiac function in an SNX–induced mouse model of CKD.²⁶ Indeed, systolic dysfunction, cardiac hypertrophy, and left ventricular dilation are present in patients with ESRD; only 16% of patients new to dialysis show normal cardiac findings on echocardiography.^27,28 The observed cardiac dysfunction and LVH in CKD mice are very likely caused by the significantly higher afterload (MABP increase of 14 mmHg). This is in line with a clinical study showing that patients on dialysis have a 48% higher risk of LVH with each increase of 10 mmHg in BP.²⁹ Hypertension is strongly related to the increased incidence of cardiovascular events in patients with stage 2 or 3 CKD.³⁰ These structural and functional alterations of heart associated with hypertension may contribute to the increased risk of cardiac death in patients with renal failure.^28,31 Thus, tight BP control attenuating the hypertensive heart disease (first hit) in patients with CKD might be crucial to prevent the underlying predisposition to second insults, such as sepsis.

Notably, we report here, for the first time, that the presence of CKD increases the severity of LPS–induced cardiac dysfunction using a two–hit animal model that consists of preexisting CKD followed by LPS injection. This is in agreement with the clinical findings that the preexisting CKD worsens outcome in patients with infection or sepsis.^16,32 We have recently reported that CLP/sepsis does not cause a significant cardiac (and indeed, multiple organ) dysfunction in young mice when these animals are treated with fluids and antibiotics, whereas older animals (8 months old) do develop cardiac (multiple organ) dysfunction, despite fluid resuscitation and antibiotics.^18,22 We show here that young mice with CKD do develop a profound cardiac (systolic) dysfunction in response to CLP, which is similar to the cardiac dysfunction in aged mice with CLP. Like CKD, ageing is associated with a mild systemic inflammation characterized by elevated plasma concentrations of IL-6, IL-1β, and TNF³³; this proinflammatory phenotype in ageing (or CKD) may be secondary to (1) the observed activation of NF-κB, which is one of the signatures of ageing,³³ or (2) impaired excretion of cytokines by the kidneys caused by decreased renal function (because of a reduced number of functional glomeruli and lower GFR).³⁴ Indeed, 24-month-old mice exhibit systemic inflammation as well as an impairment in renal function (data not shown).

NF-κB is one of the most important proinflammatory transcription factors, consisting of heterodimer subunits p50 and p65.²⁰ CKD caused cardiac phosphorylation of Ser^176/180 on IKKα/β, indicating IKK activation, which in turn, led to phosphorylation of IκBα and activation of NF-κB. Additionally, phosphorylation of IκBα can be induced by the exposure to proinflammatory cytokines, such as IL-1β and TNF-α.³⁵ Indeed, plasma proinflammatory cytokine levels were increased in CKD mice, paralleled by the increased cardiac phosphorylation of IκBα. The cardiac activation of NF-κB in CKD mice may also be attributable to the hypertensive state. NF-κB is significantly activated in rat cardiomyocytes subjected to cyclic mechanical stretch, which mimics some aspects of the pathophysiologic changes associated with hypertension in cardiac myocytes.³⁶ It is possible that the activation of NF-κB has (at least in part) contributed to the cardiomyopathy through induction of expression of its target gene iNOS. Cardiac activation of NF-κB and the subsequent iNOS expression contribute to sepsis–related impaired left ventricular function.^18,37,38 Indeed, in this study, nuclear translocation of p65 and iNOS expression were augmented in hearts of CKD/sepsis mice, and this was associated with a worsened cardiac dysfunction. Because neither low-dose LPS nor CLP significantly affected any of the above signaling pathways in mice without CKD, it is likely that the baseline cardiac activation of NF-κB during CKD acts as the prime driver of the observed excessive activation of NF-κB (and expression of NF-κB–dependent genes) and the associated cardiac dysfunction in CKD/sepsis.

In addition to inducing iNOS expression, NF-κB activation also leads to a pronounced increase in other proinflammatory cytokines.³⁹ Here, we report a dramatic increase in plasma levels of TNF-α, IL-1β, IL-6, and IL-10 in CKD mice with CLP; >70% of inflammatory cytokines are excreted by the kidney,⁴⁰ and the half-lives of TNF-α, IL-6, and IL-10 are two- to threefold prolonged in CKD mice compared with normal mice.⁴⁰ Therefore, impaired renal function resulting in a prolonged half-life of cytokines in CKD mice may amplify systemic inflammation, which in turn, may contribute to the excessive cardiac dysfunction and lung inflammation in CKD mice with sepsis.^41,42 The augmented lung inflammation in CKD mice subjected to sepsis reported in this study is in line with a number of epidemiologic studies showing that preexisting CKD predisposes patients with pneumonia to higher mortality rates.^43–45

Having found the significant roles of phosphorylation of IKKα/β and the subsequent activation of NF-κB in the augmented cardiac dysfunction induced by sepsis/endotoxemia in CKD mice, we have then investigated the role of the selective inhibition of IKK complex in vivo in CKD mice that underwent CLP or LPS administration. The treatment protocol for IKK 16 used in this study reduces systemic inflammation and organ injury in mice with sepsis without CKD.²² We found, for the first time, that a single dose of IKK 16 started 1 hour after CLP or LPS administration attenuated sepsis–induced cardiac dysfunction in CKD mice and corresponded to significant attenuated cardiac activation of NF-κB and iNOS expression. Additionally, systemic inflammatory cytokine levels in CKD/CLP or CKD/LPS mice were reduced by IKK 16, presumably by inhibiting the production of inflammatory cytokines mediated by NF-κB activation and their release into plasma.²¹ The attenuated lung inflammation with IKK 16 treatment in CKD/CLP or CKD/LPS mice was in line with previous studies, which showed therapeutic benefits of IKK 16 on sepsis–induced lung inflammation in normal mice²² and ventilation–induced lung injury.⁴⁶

Sustained high levels of activation of the phosphoinositide 3-kinases/Akt and the ERK1/2 pathways have been involved in cardiomyocyte growth and the development of cardiac hypertrophy.⁴⁷ In this study, the cardiac phosphorylation of Akt and ERK1/2 may contribute to the CKD–associated cardiac hypertrophy and cardiomyopathy. Similar to our results, the ERK1/2 pathway was also activated in rat hearts with adenine-induced CKD.⁴⁸ The activation of Akt and ERK1/2 was reduced by the administration of IKK 16 in septic CKD animals, presumably through the downregulation of NF-κB activation and the decreased expression of inflammatory cytokines, such as TNF-α.^49,50 In turn, downregulated Akt and ERK1/2 phosphorylation may lead to less NF-κB activation, decreasing cytokine production and thus, forming a feed-forward mechanism and further reducing the inflammatory reaction.^49,51

We have discovered that preexisting CKD augments the cardiac dysfunction caused by sepsis/endotoxemia. CKD alone resulted in moderate systemic inflammation and activation of NF-κB (and iNOS expression) in the heart, whereas sepsis/endotoxemia (second hit) in animals with preexisting CKD resulted in a dramatic rise in a number of proinflammatory cytokines (in the plasma) as well as a dramatic increase in the activation of NF-κB (and iNOS expression) in the heart. Most notably, selective inhibition of IKK (by administration of IKK 16 after the onset of sepsis/endotoxemia) abolished the systemic inflammation and cardiac dysfunction caused by sepsis/endotoxemia in animals with CKD. Thus, inhibition of IKK may be useful to treat the excessive inflammation and systolic cardiac dysfunction associated with sepsis in patients with CKD.

Concise Methods

Additional details on the methods are provided in Supplemental Material.

Animals

The local Animal Use and Care Committee approved animal experiments in accordance with the derivatives of both the Home Office Guidance on the Operation of Animals (Scientific Procedures) Act 1986 and the Guide for the Care and Use of Laboratory Animals of the National Research Council.⁵² This study was carried out on 117 4- to 6-week-old male C57BL/6 mice (Charles River Laboratories, Wilmington, MA) receiving a standard diet and water ad libitum.

Animal Models of SNX

Mice were subjected to a two-stage SNX. We followed the original SNX protocol introduced by Gagnon and Gallimore⁵³ with slight modifications as described in Supplemental Material. Mice subjected to sham operations were operated on without removing the kidney.

Model of LPS–Induced Organ Dysfunctions

Mice with and without CKD (CKD sham) received intraperitoneal injections of low-dose LPS (2 mg/kg) or its vehicle (PBS). Sham-treated mice were not subjected to LPS but otherwise treated the same way.

Model of Polymicrobial Sepsis Caused by CLP

Polymicrobial sepsis was induced by CLP (18-gauge needle; double puncture) in mice. Mice received volume resuscitation and antibiotic and analgesic therapy.^54,55 The detailed CLP procedure is described in Supplemental Material. Sham-operated mice were not subjected to ligation or perforation of cecum but were otherwise treated the same way. One hour after CLP, CKD mice were treated with either IKK 16 (1 mg/kg intravenously) or vehicle (2% DMSO).

Radiotelemetric Recording of Hemodynamics and Activity In Vivo

BP was recorded in conscious, freely moving mice using radiotelemetric transmitters (TA11PA-C10; Data Sciences International) implanted into the aortic arch. After 10 days of recovery, the BP and activity were recorded for 3 hours before and 20 hours after CLP surgery. Data were acquired for 2 minutes every 15 minutes, and the average values for MABP (millimeters of mercury) and activity (arbitrary units) were calculated for every time point (Dataquest Art Acquisition System). ΔMABP and Δactivity were calculated at each time point by subtracting the reading from the average measurement during the 3-hour baseline recordings.

BP Recording in Anesthetized CKD Mice

Mice were anesthetized with 2% isoflurane delivered in 0.4 ml/min oxygen. MABP was measured via carotid artery using a fluid-filled catheter and a BP transducer (MLT1199; AD Instruments). A 10-minute baseline recording was taken, IKK 16 (1 mg/kg) was given intravenously via the jugular vein, and MABP was monitored for 1 hour.

Quantification of Organ Dysfunction/Injury

Cardiac function was assessed in mice subjected to LPS at 18 hours or CLP at 24 hours by echocardiography using a Vevo-770 Imaging System (Visual Sonics, Toronto, Canada).^55,56 Then, the experiment was terminated, and organ and blood samples were collected for quantification of organ dysfunction/injury. Details are available in Supplemental Material.

Western Blot Analyses

We analyzed the degree of phosphorylation of IKKα/β on Ser^176/180, IκBα on Ser^32/36, Akt on Ser⁴⁷³, and ERK1/2; the nuclear translocation of the p65 subunit of NF-κΒ; and the expression of iNOS. Semiquantitative Western blot analyses were carried out in mouse heart tissues as described previously⁵⁷ and outlined in Supplemental Material.

Determination of MPO Activity in Lung Tissue

MPO was extracted from the tissue as described by Barone et al.⁵⁸ with slight modifications. MPO activity used as a marker for neutrophil accumulation in tissues was determined as previously described.⁵⁹

Measurement of Cytokines

Concentrations of cytokines in culture supernatants and plasma were measured using a commercially available cytometric bead array (BD Biosciences, San Jose, CA or BioLegend, San Diego, CA) as described in the manufacturer’s instructions. Details are available in Supplemental Material.

Statistical Analyses

Values are presented as means±SEMs of n observations. Data were assessed by a one-way ANOVA followed by Bonferroni post hoc test (multiple comparison), a two-way ANOVA followed by Bonferroni post hoc test (time course and multiple comparison), unpaired t test, or Mann–Whitney U test. P<0.05 was considered to be statistically significant.

Disclosures

None.