Pro-oxidative priming but maintained cardiac function in a broad spectrum of murine models of chronic kidney disease

Abstract

Aims: Patients with chronic kidney disease (CKD) have an increased risk of cardiovascular events and exhibit myocardial changes including left ventricular (LV) hypertrophy and fibrosis, overall referred to as 'uremic cardiomyopathy'. Although different CKD animal models have been studied for cardiac effects, lack of consistent reporting on cardiac function and pathology complicates clear comparison of these models. Therefore, this study aimed at a systematic and comprehensive comparison of cardiac function and cardiac pathophysiological characteristics in eight different CKD models and mouse strains, with a main focus on adenine-induced CKD.

Methods and results: CKD of different severity and duration was induced by subtotal nephrectomy or adenine-rich diet in various strains (C57BL/6J, C57BL/6 N, hyperlipidemic C57BL/6J ApoE^-/-, 129/Sv), followed by the analysis of kidney function and morphology, blood pressure, cardiac function, cardiac hypertrophy, fibrosis, myocardial calcification and inflammation using functional, histological and molecular techniques, including cardiac gene expression profiling supplemented by oxidative stress analysis. Intriguingly, despite uremia of variable degree, neither cardiac dysfunction, hypertrophy nor interstitial fibrosis were observed. However, already moderate CKD altered cardiac oxidative stress responses and enhanced oxidative stress markers in each mouse strain, with cardiac RNA sequencing revealing activation of oxidative stress signaling as well as anti-inflammatory feedback responses.

Conclusion: This study considerably expands the knowledge on strain- and protocol-specific differences in the field of cardiorenal research and reveals that several weeks of at least moderate experimental CKD increase oxidative stress responses in the heart in a broad spectrum of mouse models. However, this was insufficient to induce relevant systolic or diastolic dysfunction, suggesting that additional "hits" are required to induce uremic cardiomyopathy.

Translational perspective: Patients with chronic kidney disease (CKD) have an increased risk of cardiovascular adverse events and exhibit myocardial changes, overall referred to as 'uremic cardiomyopathy'. We revealed that CKD increases cardiac oxidative stress responses in the heart. Nonetheless, several weeks of at least moderate experimental CKD do not necessarily trigger cardiac dysfunction and remodeling, suggesting that additional "hits" are required to induce uremic cardiomyopathy in the clinical setting. Whether the altered cardiac oxidative stress balance in CKD may increase the risk and extent of cardiovascular damage upon additional cardiovascular risk factors and/or events will be addressed in future studies.

Keywords: Animal model; Cardiac remodeling; Cardiomyopathy; Chronic kidney disease; Oxidative stress.

Fig. 1

Schematic overview of studied mouse models and results summary. CKD was induced in C57BL/6J, hyperlipidemic C57BL/6J ApoE^−/−, C57BL/6 N or 129/Sv, by 5/6 nephrectomy (5/6 Nx) or by feeding an adenine-supplemented diet (diet details in Suppl. Table 1). Although overall, no cardiac dysfunction, hypertrophy or interstitial fibrosis could be observed, several weeks of at least moderate CKD did alter oxidative stress responses in the heart and enhanced cardiac oxidative stress markers in each mouse strain. 129/Sv mice with moderate to severe CKD as well as C57BL/6 N with severe CKD also developed myocardial calcified deposits surrounded by localized fibrotic tissue. COL1 = collagen 1; CO = cardiac output; (dP/dT)max: maximum rate of left ventricular pressure rise over time (mmHg/s); EF = ejection fraction; HFD = high-fat diet; HO = heme oxygenase; ICAM = intercellular adhesion molecule; 8-OHdG = 8-hydroxy-2-deoxyguanosine; SMA = smooth muscle actin.

Fig. 2

MODEL 3–5/6 Nx in C57BL/6 N mice. A) Experimental timeline. BL = baseline. BP = blood pressure. B) Body-weight curve (Sham n = 4; 5/6 Nx n = 4). C) Plasma creatinine and urea. D) Representative images of kidney AFOG and PAS staining. E) Ejection fraction and cardiac output via echocardiography. F) Systolic and diastolic blood pressure at the endpoint. G) Analysis of Anp/Bnp gene expression in cardiac tissue (normalized to Gapdh) and cardiomyocyte diameter in WGA-stained cardiac sections. H) Quantification and representative images of histological Sirius Red staining of cardiac sections. I) Quantitative PCR on cardiac tissue for markers of fibrosis (Tgfb1, Col1a1, Col3a1) and inflammation (Tnf, Icam1, Ccl2), normalized to Hprt1 and GusB. B–I) Unless otherwise indicated, endpoint analyses were performed and data are presented as means ± SD or dot plots. *p < 0.05 comparing nephrectomy to sham animals using mixed-effects analysis with matching values and Sidak's post-test (B), two-tailed t-test (parametric data; with Welch's correction in case of non-equal SDs) or Mann-Whitney test (non-parametric data) (C–J). (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

Fig. 3

MODEL 5 - High-dose Adenine diet in C57BL/6J ApoE^−/− mice. A) Experimental timeline. HFD = high fat diet, BP = blood pressure, EP = endpoint. B) Body-weight curve (Sham n = 9; Adenine n = 10). C) Serum creatinine and urea, measured at 2, 4 and 6 weeks. D) Representative images of kidney AFOG and PAS staining (PAS: asterisk indicates tubular injury; arrow indicates infiltrating cells). E) Quantitative analysis of COL1 and αSMA as fibrosis markers in kidney tissue via western blot, normalized to loading control GAPDH. F) Invasive heart function analysis at endpoint by Millar catheter: (dP/dT) max and min. G) Left ventricular end diastolic pressure (LVEDP). H) Quantitative analysis of cytosolic (SOD1, PRX2, CAT) and mitochondrial (SOD2, PRX3) antioxidative enzyme expression as well as oxidative stress markers (NOX2, HO-1) in heart tissue lysates via western blot, normalized to loading control (GAPDH or γTUB). I) 8-OHdG immunostaining in heart sections. B-I) Shown are means ± SD or dot plots. *p < 0.05, **p < 0.01, p***<0.001 comparing adenine to sham animals using two-way ANOVA and Sidak's post-test (B–C, F-G), two-tailed t-test (parametric data; with Welch's correction in case of non-equal SDs) or Mann-Whitney test (non-parametric data) (E, H–I).

Fig. 4

MODEL 7 - Low-dose Adenine diet in C57BL/6 N mice. A) Experimental timeline. BL = baseline. B) Body-weight curve (Sham n = 4; Adenine n = 4). C) Plasma creatinine and urea measured at 6, 9, 12, 15 and 16 weeks (missing values due to insufficient blood collection and no 9 weeks data for urea due to limited blood availability). D) Quantitative analysis of COL1 and αSMA as fibrosis markers in kidney tissue via western blot, normalized to loading control GAPDH, and representative images of kidney AFOG and PAS staining (PAS: asterisk indicates tubular injury; arrow indicates infiltrating cells). E) Ejection fraction and cardiac output via echocardiography. F) Cardiac analysis of Anp/Bnp expression (normalized to Gapdh) and cardiomyocyte diameter in WGA-stained cardiac sections. One Anp outlier excluded in Adenine group based on the Grubb's test. G) Quantitative PCR on cardiac tissue for markers of fibrosis (Tgfb1, Col1a1, Col3a1) and inflammation (Tnf, Icam1, Ccl2), normalized to Hprt1 and GusB. H) Ratiometric (405 nm/488 nm) mitochondrial redox analysis in the heart (roGFP2-Orp1) at 16 weeks. I) Representative images of ratiometric (405 nm/488 nm) mitochondrial redox analysis in heart sections. J) Quantitative analysis of cytosolic (SOD1, PRX2, CAT) and mitochondrial (SOD2, PRX3) antioxidative enzyme expression in heart tissue lysates via western blot, normalized to loading control (GAPDH or γTUB). K) Quantification of oxidative stress markers via western blot (NOX2, HO-1), normalized to loading control (GAPDH or γTUB) and cardiac 8-OHdG immunostaining. B–K) Unless otherwise indicated, endpoint analyses were performed and data are presented as means ± SD or dot plots. For mice sacrificed prematurely at day 109 or day 112, kidney and heart were still collected for organ analysis of the endpoint (day 113). *p < 0.05, **p < 0.01 comparing adenine to sham animals using mixed-effects analysis with matching values and Sidak's post-test (B–C; no statistical evaluation for groups with n = 2 only), two-tailed t-test (parametric data; with Welch's correction in case of non-equal SDs) or Mann-Whitney test (non-parametric data) (D–K).

Fig. 5

MODEL 8 - Low-dose Adenine diet in 129/Sv mice. A) Experimental timeline. BL = baseline. B) Body-weight curve (Sham n = 4; Adenine n = 4). C) Plasma creatinine and urea. D) Representative images of kidney AFOG and PAS staining (PAS: asterisk indicates tubular injury; arrow indicates infiltrating cells). E) Ejection fraction and cardiac output via echocardiography. F) Cardiac analysis of Anp/Bnp expression (normalized to Gapdh) and cardiomyocyte diameter in WGA-stained cardiac sections. G) Representative images of adenine-induced myocardial deposits: macroscopic view, hematoxylin-eosin, Sirius Red, von Kossa and Alizarin Red stainings. H) Quantitative PCR on cardiac tissue of markers for fibrosis (Tgfb1, Col1a1, Col3a1) and inflammation (Tnf, Icam1, Ccl2), normalized to Hprt1 and GusB. I) Quantitative analysis of cytosolic (SOD1, PRX2, CAT) and mitochondrial (SOD2, PRX3) antioxidative enzyme expression in heart tissue lysates via western blot, normalized to loading control (GAPDH or γTUB). J) Quantitative analysis of oxidative stress markers (NOX2, HO-1) in heart tissue lysates via western blot, normalized to corresponding loading controls (GAPDH or γTUB). K) Quantification of 8-OHdG immunostaining in heart sections. B-K) Unless otherwise indicated, endpoint analyses were performed and data are presented as means ± SD or dot plots. *p < 0.05, **p < 0.01 comparing adenine to sham animals using mixed-effects analysis with matching values and Sidak's post-test (B–C), two-tailed t-test (parametric data; with Welch's correction in case of non-equal SDs) or Mann-Whitney test (non-parametric data) (E–K). (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

Fig. 6

RNA-sequencing reveals pathology-relevant molecular changes in the heart in 129/Sv mice with CKD. RNA-sequencing was performed on cardiac tissue of 129/Sv mice in CKD versus control conditions (Model 8). A) Volcano plot depicting log₂(fold change) and –log₁₀(p_adjusted) of genes, comparing the adenine-treated group to controls. Differentially expressed genes (DEGs) with p_adjusted < 0.05 and log₂(fold change) > 1 or < −1 were considered as significant for further analysis, and are marked in green (downregulated) or red (upregulated); ten significantly up- and downregulated genes of interest were highlighted (green: downregulated genes; red: upregulated genes). B) Representation of enriched and pathology-relevant gene ontology (GO) terms based on the significant DEGs. Green bars indicate downregulated and red bars upregulated pathways. Selected significant DEGs associated with these GO terms are indicated. ECM = extracellular matrix; ROS = reactive oxygen species. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

Fig. 7

Molecular adaptation of the heart in CKD. 129/Sv mice on adenine-induced CKD revealed an increased cardiac oxidative stress response characterized by an increase in ROS-inducing mediators and a decrease in cardioprotective mediators, as identified by RNA-sequencing. On the other hand, also protective anti-inflammatory feedback responses were detected. Overall, cardiac levels of the oxidative stress markers HO-1 and 8-OHdG were increased. Since oxidative stress signaling can both offer protective as well as detrimental effects, future studies should address whether this oxidative signature provides a certain degree of cardioprotection or in fact increases the risk of enhanced cardiovascular damage upon additional cardiovascular risk factors and/or events.