Cardiac dysfunction in C57Bl/6J mice with chronic kidney disease shows sex-specific effects: Comparison of dietary adenine and 5/6 nephrectomy models

Background

Cardiovascular disease (CVD) is the leading cause of death among patients with chronic kidney disease (CKD). Rodent models are widely used to study uremic CVD pathophysiology. We compared cardiac function parameters in male and female animals from 2 established mouse CKD models and examined associations between gut-derived uremic toxins and echocardiogram findings.

Methods

Male and female adult C57Bl/6J mice were randomly assigned to control, adenine-induced CKD and 5/6 nephrectomy CKD groups. Echocardiography was performed on all mice at age 17 weeks (5 weeks after CKD induction). Serum creatinine, cystatin C and gut-derived uremic toxins were analyzed at study termination, and RNA sequencing of left ventricle tissue was performed and analyzed.

Results

Markers of kidney dysfunction were elevated in both CKD models. The gut-derived uremic toxin indoxyl sulfate (IS) was increased in both CKD models, while trimethylamine N-oxide (TMAO) was increased in adenine CKD mice and p-cresyl sulfate (pCS) in nephrectomy animals. Left ventricular volume was increased in nephrectomy animals. Cardiac output (CO) was decreased in male CKD animals from both models compared to controls, and ejection fraction (EF) was decreased in male 5/6 nephrectomy mice. Female controls had lower stroke volume (SV) and CO than male counterparts. No significant changes in CO or EF were observed in female CKD groups. Higher serum pCS and IS positively correlated with increased left ventricular volume; higher pCS inversely correlated with EF in females but not males. Transcriptomics of cardiac tissue revealed sex-based variations in matrix metalloproteinase and mitochondrial pathways associated with cardiac dysfunction.

Conclusions

Our work highlights sex differences in cardiac function and serum chemistries in two established preclinical CKD models. Gut-derived uremic toxins may significantly impact cardiorenal pathophysiology in female but not male CKD animals.

Methods
Animal models

All animal studies were approved by the Animal Research Committee, University of California, Irvine (UCI) under protocol AUP-21-157. Animals were maintained at the UCI vivarium in accordance with the policies instituted by the American Association for Accreditation of Laboratory Animal Care. Male and female C57Bl/6J mice aged between 8-10 weeks were purchased from the Jackson Laboratory (Bar Harbor, ME) and were randomly assigned to control vs adenine-induced CKD. Mice in adenine groups were fed a 0.2% adenine diet for 18 days, followed by 2 weeks on regular chow and then a second 1-week period on adenine diet to maintain CKD . Age-matched male and female 5/6 nephrectomized mice were purchased from the Jackson Laboratory. The initiation timepoint of uremia was considered to be after 18 days of adenine diet (adenine-CKD model) or 14 days post-surgery (5/6 nephrectomy model). Echocardiography was performed at 5 weeks after uremia induction at age 18 weeks (n=8 per group). Tail blood pressure was measured prior to study termination via tail-cuff plethysmography which included 10 habituation cycles followed by 10 measurements, with automated exclusion of outlier values (CODA-S2 multi-channel, Kent Scientific). Body weight was monitored weekly to assess the overall health of the mice and to track the progression of kidney disease, and body weight at termination is presented in Table 1. At study termination, blood was collected via cardiac puncture and processed for serum collection. Heart tissue was collected for RNA analysis.

Echocardiography

Echocardiography was performed at 5 weeks after uremia induction, at age 18 weeks . After anesthesia induction with isoflurane, animals were maintained under inhaled 1.5% isoflurane and 95% O2 to allow for spontaneous breathing. Animals were imaged in a left lateral decubitus position and warming pads were used to maintain normothermia. Images of the horizontal long-axis view and short-axis view of the heart were collected in both M-mode and B-mode [19] using a Vevo 3100 ultrasound system with 25-mHz scan head (VisualSonics, Toronto, ON, Canada). The “LV trace” feature was used to calculate LV structural and functional parameters from the LV parasternal short-axis M-mode view, which was recorded at the level of two papillary muscles. LV posterior wall thickness during diastole was measured on B-mode long-axis view. Vevo analysis software (VevoLab, VisualSonics) was used to conduct all measurements and calculations. Vevo software analysis was performed by a research team member blinded to study groups (H.W.).

Serum chemistries

Serum creatinine collected at study termination was analyzed via liquid chromatography-mass spectrometry at the O’Brien Center, University of Alabama at Birmingham (Birmingham, AL). Cystatin C was measured using the Invitrogen Cystatin C (CST3) Mouse ELISA Kit (EMCST3, Fisher Scientific). Total-form serum levels of the gut-derived uremic toxins p-cresyl sulfate, indoxyl sulfate and TMAO were measured using liquid chromatography with tandem mass spectrometry (HPLC-MS/MS) based on previously described methods. A 25 mL aliquot of serum was treated with 200 uL of acetonitrile with 0.1% formic acid and internal standard for protein precipitation. The mixture was vortexed and centrifuged, and evaporated to dryness. The dried extract was reconstituted with 100 uL of 25% acetonitrile. Standards and prepared samples were injected (10 ml) into the HPLC-MS/MS instrument, Waters Quattro Premier XE equipped with UPLC. For indoxyl sulfate and p-cresyl sulfate, hydrochlorothiazide was used as the internal standard and analysis was performed in negative ionization mode. For TMAO, salbutamol was used as the internal standard and analysis was performed in positive ionization mode.

RNA sequencing

Total RNA was isolated from left ventricular heart tissue using TRIzol® reagent (Life Technologies, Carlsbad, CA, USA), followed by an additional DNase I digestion to remove genomic DNA contamination. RNA samples were submitted to the UC Irvine Genomics Research and Technology Hub (GRTH) for sequencing with the NovaSeq 6000 platform S4 (Illumina, San Diego CA). RNA quality was verified using the Agilent Bioanalyzer Nano RNA chip (Agilent Technologies, Santa Clara CA) and Nanodrop (Thermo Fisher Scientific, Waltham MA). Samplesin the current study showed A260/A280 absorbance ratios in the range 1.928-2.236. Library construction was performed according to the Illumina TruSeq mRNA stranded protocol using the Apollo 324 library prep system (Takara Bio, Mountain View CA). The input quantity for total RNA was 1µg and mRNA was enriched using oligo-dT magnetic beads. The enriched mRNA was chemically fragmented for three minutes. First-strand synthesis used random primers and reverse transcriptase to make cDNA. After second-strand synthesis, the cDNA was cleaned using AMPure XP beads, followed by end repair and 3’ adenylation. Illumina dual-indexed adapters were ligated on the ends and the adapter-ligated fragments were enriched by nine cycles of PCR. The resulting libraries were validated by qPCR (KAPA Library quantification kit, Kapa Biosystems, Wilmington MA) and sized with the Agilent Bioanalyzer DNA high-sensitivity chip. Library concentrations were normalized and then multiplexed together. The multiplexed libraries were sequenced using paired-end with read length 150 base pairs. Fastq files obtained from sequencing were inspected via FastQC then aligned to the current mouse genome build (mm10) using STAR. Counts matrices were then generated from aligned BAM files using summarizeOverlaps (Genomicfeatures) and differential expression was performed using DESeq2 with false discovery rate (FDR) adjusted P-value <0.001. All plots were generated in R. Pathways derived from differentially expressed genes (DEGs) were identified using Gene Set Enrichment Analyses (GSEA) based on erichR focusing queries on Reactome, Gene Ontology and KEGG databases.

Statistical analysis

Data are presented as the mean ± standard error of the mean (SEM). Outliers were excluded using the ROUT test with Q=1%. Two-way analysis of variance (ANOVA) with Tukey’s multiple comparisons test were performed using GraphPad Prism 8.4 (GraphPad Software, San Diego, CA, USA). Šídák's multiple comparisons test was used to determine sex differences. Blood chemistries did not follow a Gaussian distribution per the Kolmogorov-Smirnov normality test, therefore nonparametric Spearman correlation analysis was used to analyze associations between toxin levels and echocardiographic parameters. Linear regression analysis was computed including regression equation, slope p-values, and R2. P-values <0.05 were considered statistically significant.