Background
Chronic reno-cardiac syndrome (RCS), a branch of the general cardio renal syndrome in which impaired renal function inflicts consequential damage on to the cardiac vasculature (defined by Ronco et al. [
1]) is prevalent in the end stage renal disease (ESRD) population having a 20 fold higher incidence of cardiovascular mortality (encompassing the clinical scenarios of myocardial infarction, sudden death, arrhythmia and cardiomyopathy) compared to the population as a whole [
2,
3]. Approximately 50% of mortality in ESRD patients is as a result of cardiovascular events [
4,
5]. These factors highlight the need for the continued improvement of therapeutic strategies to combat RCS.
The employment of animal models as tools to explore disease pathologies has greatly enhanced research especially with the development of technology to genetically modify strains. The most successful animal model of RCS is the sub-total nephrectomy model in rats where a portion (5/6ths) of the kidney mass is surgically removed or negated [
6]. The resulting renal ischaemia initiates stimulation of the sympathetic nervous system, triggering the renin–angiotensin–aldosterone system and inhibiting nitric oxide synthesis eventually causing hypertension, left ventricular hypertrophy (LVH) and progressive left ventricular dilatation [
7]. Fibroblast activation leads to eventual cardiac hypertrophy and fibrosis [
8,
9]. Unfortunately in mice, this model is much less effective giving highly variable results [
10]. This is further compounded by the fact that the C57BL/ 6 mouse strain, the preferred strain for genetic modification [
11] has been shown to be resistant to developing CKD by subtotal nephrectomy [
12].
Another model of CKD is a chemical method involving the addition of adenine to the diet. The peculiar metabolism of adenine compared to other purines was first described in rats by Yokozawa et al. [
13] and its nephrotoxic effects were subsequently employed in the creation of a highly reproducible model of CKD in rats. However, adenine-induced CKD results in a rapid impairment of renal function and high (early) mortality, which prevents the development of RCS. In mice, the adenine model has been less successful due to their distaste of adenine causing substantial and rapid weight loss leading to mortality after 4 weeks [
14]. The addition of casein into the diet masks the taste of the adenine and prolongs the model to 8 weeks [
14] however this is still not long enough for cardiovascular effects to mature.
The aim of this study was to establish a non-surgical and, therefore, more replicable model of RCS in the C57BL/ 6 mouse strain. To achieve this we further manipulated the adenine diet protocol by reducing the volume of adenine in the diet to 0.15% thus, concealing the taste of the adenine (without the need for addition of casein), to the point that mice ingested it without weight loss and corresponding morbidity. In this way the diet regimen was extended to 20 weeks causing the development of CKD and subsequent RCS.
Discussion
Translational medicine requires the ability to investigate the in vivo expression of pathological genes in RCS thus necessitating the development of a reliable model of CKD in the C57BL/ 6 mouse the most common background strain for transgenic and knockout breeds.
Constraining factors in currently available murine models of RCS include high variability due to limitations of surgical skill and high mortality due to the intolerance of mice to such acute surgical intervention [
21‐
25]. This is further confounded by the apparent resistance of the C57BL/ 6 strain to surgically induced uraemia [
12].
The induction of CKD by the administration of a diet containing high adenine content has been successfully employed in rat studies of renal failure [
26]. The metabolic product of adenine is 2,8-dihydroxyadenine which forms a precipitate of crystals in the proximal tubule specifically in the microvilli and apical epithelial region causing inhibition of tubular function and resulting in hyperphosphatemia, secondary hyperparathyroidism, bone disease and vascular calcification [
27,
28].
A model based on diet administration as opposed to surgical procedure would obviously alleviate operator variability however the use of adenine diet to induce CKD in mice has been unsuccessful due to their aversion to the taste of adenine resulting in malnutrition based morbidity and mortality. Although manipulations of the adenine regimen have met with some success, the requirement of prolonging the model long enough to induce RCS has not yet been achieved. To meet this necessity we have investigated the effects of administration of a less detectable low dose of adenine diet over an extended period of 20 weeks.
Our results showed that the mice tolerated the diet well with no weight loss throughout the experimental period. Unlike surgical models of CKD there was no premature mortality and endpoint serum and urine biochemistry showed that all animals developed a similar degree of CKD.
Interestingly, unlike other adenine regimens, serum creatinine was significantly increased in CKD mice indicating no sign of malnourishment and also unlike other adenine regimens, the fact that the diet was given at a constant dose throughout meant that there was no requirement to monitor serum parameters during the course of the experiment in order to adjust the dosage accordingly.
Echocardiography suggested a small, but significant, impairment in systolic function (EF) in mice with adenine diet. However echocardiography was unable to detect an increased interventricular septum thickness in the adenine group but rather an increase in left ventricular dimensions (left ventricular internal diastolic dimension or left ventricular end diastolic volume, LVID). This taken together with a significantly greater heart weight-to body weight ratio (a surrogate marker for myocardial hypertrophy [
29]), could imply development of an eccentric hypertrophy of the left ventricle (dilation of left ventricular chamber coupled with mild to moderate increased wall thickness [
30]). Up to 50% of patients with heart failure have diastolic dysfunction with relatively well-preserved EF, which is termed “heart failure with preserved ejection fraction (HFpEF)” [
31]. Due to the mild decrease in EF in mice with adenine diet, it is possible that this model can be characterised as HFpEF, however future study of diastolic function in this model is needed to confirm this.
We did not measure blood pressure in these mice however we have previously reported a lower degree of hypertension in adenine induced uremic animals when compared to sub-total nephrectomised animals despite a greater level of renal failure in the former [
32].
RCS was further confirmed by a significantly higher degree of cardiac phosphorylation of Akt suggesting upregulation of cell survival mechanisms in response to injury [
33], and evidence of continuing detrimental effects on the heart [
34,
35]. H and E, CD45 and F4/80 staining reinforced evidence of inflammation. There was also an increased expression of the fibrotic biomarkers fibronectin, α-SMA, sirius red and collagens 1 and 3 in the intermuscular spaces of the hearts indicating increased myofibroblast formation [
36,
37] possibly due to increased pre or afterload hypertrophy [
38].
Fibrosis in the kidneys was confirmed with the same biomarkers and by a higher expression of renal total Akt, an important regulator of epithelial-mesenchymal transition of tubular epithelial cells into a myofibroblast phenotype, another indicator of compensatory renal hypertrophy [
39‐
41].
Overall these results imply that our method has been effective in the induction of CKD and RCS in the C57BL/ 6 mouse strain. It’s main advantage over surgical models being its reduction in disparity between operators thus lowering animal volume required per experiment and also eliminating the necessity for surgical expertise and equipment. Its main advantage over other adenine based regimens being its simplicity and longevity through decreased morbidity.
We acknowledge that other nonsurgical models of RCS are already in existence and deserve comment. The administration of carbon tetrachloride is an established method of causing simultaneous cardiac and renal dysfunction and has been successfully employed in the C57BL/ 6 strain [
42]. However this method is known to initiate undesirable injury in other organs [
43]. This is also the case with other forms of chemical nephrectomy where the physiological effects of agents such as cisplatin, uranyl nitrate or adriamycin on the whole body system are impossible to supervise [
44‐
46].
Obviously we recognise the fact that a chemical nephrectomy model such as adenine administration has disadvantages. As mentioned by Jia et al. [
14], an adenine model is based on adenine metabolites causing tubular toxicity and therefore a tubular-interstitial disease rather than glomerular scarring secondary to vascular damage which is the most familiar source of human CKD. However, even in human CKD, tubular interstitial fibrosis is the best predictor of renal failure irrespective of the aetiology of the renal insult. Accepting these facts we realise the limitations of our method of trying to physiologically mimic the complete RCS scenario however such limitations apply to many established disease models.
Acknowledgements
Not applicable.