Subtotal (5/6) nephrectomy is a widely used, reliable model of CKD in rodents. It mimics the mechanism of CKD secondary to nephron loss in humans, characterized by progressive glomerulosclerosis and tubulointerstitial fibrosis[87]. Subtotal nephrectomy can be performed by either ligation of the polar branches of the renal artery (ligation model) or by excision of the poles (ablation model), followed by contralateral nephrectomy. Although the ligation model is associated with more severe hypertension and proteinuria[4], the extent of resulting nephron loss is fairly unpredictable, causing a significant variation in the severity of CKD. However, the ablation model seems to be more reproducible with satisfactory inter-individual variation[48], thus we chose this method for CKD induction.
To confirm the development of CKD, urine creatinine and protein levels, as well as serum carbamide, creatinine, lipid and ion levels were measured 8 and 9 weeks after nephrectomy, respectively. The significant increase in serum carbamide and creatinine levels along with the decrease in urine creatinine concentration confirm the deterioration of renal function in the nephrectomized animals. Based on the study of Ormrod and Miller, the average serum creatinine level of 19.98 mmol/L corresponds to a moderate level of uremia[57].
Results of the blood test and urinalysis underline the presence of CKD since we detected a significant elevation of serum potassium and phosphate as well as calcium and magnesium. Potassium and phosphate retention are well-known consequences of CKD[30, 82], whereas the elevation of serum calcium concentration could result from the development of secondary hyperparathyroidism. In fact, secondary hyperparathyroidism is known to develop from stage 2 CKD in humans, and a four-fold increase in parathyroid hormone (PTH) level can be observed when end-stage kidney disease develops[19]. Besides, the excretion of magnesium decreases with the advancement of CKD, typically causing hypermagnesemia from stage 4[54]. Thus, the increase in magnesium concentration might be also due to the impaired renal excretion in the CKD group.
Considering the controversial results available in the literature on the effect of CKD on anxiety-related behavior, first, we conducted a battery of behavioral tests, known to be sensitive for detecting anxiety-related behaviors. The EPM test is probably the most frequently used behavioral test for the assessment of anxiety-like behavior in rodents[80]. It is an ethologically-based, so-called approach-avoidance test, based on the innate conflict between the rodents’ drive to explore the novel environment and their fear of exposed areas[10]. The time spent in the open arms and the number of entries into the open arms negatively correlate with anxiety-like behavior[20]. In our study, the CKD group spent less time in the central area and more time in the closed arms of the EPM than the sham-operated animals. The data on the ambulatory activity of the animals is somewhat contradictory since nephrectomy seems to induce a decrease in the total entries into arms, whereas it increases the total time spent in arms. Nevertheless, taking these results together with that of the OF experiments, we conclude that it’s unlikely that 5/6 nephrectomy affects locomotor activity. Therefore, the increased closed-arm time indicates an avoidance of open arms, which is a sign of anxiety-like behavior82. It should be noted, however, that no significant difference was detected in the open-arm entries and open-arm time, so the results should be interpreted with caution.
Since its development in 1934 as a test for the assessment of emotionality in rodents, the OF test has become a widely used model in preclinical tests of anxiolytic drugs. In this ethologically based test, the anxious animals tend to exhibit wall-hugging behavior (thigmotaxis), avoiding the central part of the apparatus[60]. 9 weeks after the nephrectomy, a significant decrease in the time spent in the central zone and the distance covered in the central zone was observed in the CKD group, which can be interpreted as anxiety-like behavior. Furthermore, the CKD group exhibited a decrease in rearing behavior, affecting the number of supported and unsupported rearings equally. Rearing – i. e. temporarily standing on the hind legs – is generally considered an exploratory behavior[70]. However, its relation to anxiety is rather controversial, as some studies have associated increased anxiety with increased rearing[5, 29], while others have found an increase in rearing after treatment with anxiolytic drugs[17, 25]. Recent studies have suggested that only unsupported rears (without leaning to a wall) correlate with emotionality in rodents84 and can be considered a hippocampus-related exploratory behavior[46]. In addition, an increased tendency in the immobility time was observed in the CKD group suggesting freezing behavior characteristic of anxiety. In our study, there was no change in ambulation time and ambulation distance, so the total ambulatory activity was not affected by the nephrectomy. Therefore, the reduction in rearing could more likely be attributed to the suppression of exploratory behavior, which is common in anxiety-provoking, aversive circumstances[70].
Overall, our results suggest that CKD indeed induces anxiety-like behavior. This is in accordance with the study of Chandanathil et al., who reported a decrease in peripheral square crossings and in the number of rearings both four and eight weeks after nephrectomy in rats[12]. Another group has also observed a lower number of square crossings in the OF test in moderate CKD, as well as a reduction in rearing activity in severe CKD, 1 month after nephrectomy[74].
In contrast with our observations, Tóthova et al. have not found any significant behavioral alteration in rats 3 months after 5/6 nephrectomy. 9 months after nephrectomy, however, the animals spent more time in the light compartment of the light-dark box, which suggests an anxiolytic-like effect, although the reduction of anxiety-like behavior was not confirmed by the OF test[75]. It should be noted, that most studies reported anxiety-like behavior 4–8 weeks after nephrectomy[12, 74], and in the study of Chandanathil et al., the reduction in exploratory behavior was no longer significant 12 and 16 weeks after nephrectomy[12]. In a murine model of CKD, anxiolytic-like behavior was reported 4 and 10 weeks after induction: the time spent in the light compartment of the light-dark box increased, whereas the closed arm time in the EPM test decreased. However, the reduction in closed-arm time could result from the robust increase in time spent at the central intersection of the arms, rather than from the avoidance of open arms[15]. This study has also employed a different CKD model (cortical electrocautery and nephrectomy) and a different species, which might account for some of these discrepancies.
To obtain a more comprehensive behavioral profile an MB test was also performed. To our knowledge, MB behavior has not yet been assessed in rodent CKD models. Burying behavior, i.e. the displacement of bedding material using the snout and forepaws in a concerted effort to cover a harmful or non-harmful object, is part of the rodents’ normal behavioral repertoire[7]. The MB test was originally devised to measure defensive burying as a sign of novelty-induced anxiety, but nowadays it is frequently used as an assay of perseverative, compulsive-like behavior, as well[73]. 7 weeks after nephrectomy, no difference was found in the number of buried marbles between the groups. When considering the goal-oriented interactions with marbles, however, a significant reduction in their duration and number was observed in the CKD group. As the general locomotor activity did not seem to be affected in the OF and EPM, our MB results may correlate with a reduction in explorative behavior.
Considering our findings of an anxiety-inducing effect of 5/6 nephrectomy after 7 weeks, we aimed to investigate the possible mechanism of action via assessing the presence of uremic toxins associated with anxiety as well as analyzing HPA axis activity including amygdalar expression of key anxiogenic genes.
The background of reduced exploratory behavior is not clear, but uremic toxins might be involved, as they have been implicated in the neuropsychiatric complications of CKD by several groups. In CKD patients, serum indole-3-acetic acid concentration has correlated with anxiety, depression, and the quality of sleep[35]. In rodent models of CKD, both pCS [72] and IS have been linked to anxiety-like behavior[34, 71]. IS has also reduced brain serotonin, dopamine, and norepinephrine levels in rats[34]. In our study on week 9, there was a robust increase in the concentration of both IS and pCS. It is possible therefore that a CKD-induced decrease in cerebral monoamine concentrations can be linked to reduced exploration, as fluvoxamine (a selective serotonin reuptake inhibitor) and bupropion (an atypical antidepressant) have increased exploration around the marbles in the MB test in mice[28]. Furthermore, these gut bacteria-derived, protein-bound uremic toxins (IS and pCS) have a well-established proinflammatory effect[78], which in turn has been implicated in the development of anxiety[18].
Furthermore, in our experiments, the metabolism of tryptophan has shifted from the indole and serotonin pathways to the kynurenine pathway (see Fig. 7). In CKD, multiple studies have reported an increase in KYN metabolites[6, 84], behind of which is most likely chronic inflammation, as inflammatory cytokines (such as IFN-γ, TNF-α) activate IDO, the rate-limiting enzyme of KYN synthesis[91]. KYN metabolites play a controversial role in the pathomechanism of CKD, as some of them seem to have a detrimental effect, whereas others might be considered protective[52]. For instance, 3-OH-KYN has been associated with mitochondrial dysfunction and increased ROS production[62], but it has been reported that KYNA has anti-inflammatory properties and it prevents homocysteine-induced damage to the endothelium[3, 85]. Our results showed an increase in the KYN/Trp ratio indicating an activation of IDO and/or TDO, which is in accordance with previous results[58, 33]. TDOs are mainly localized in the liver and can be activated in response to glucocorticoids. However, most of the kynurenine synthesis in inflammatory diseases is extrahepatic, as cytokines activate IDOs in the central nervous system, blood, spleen, kidneys, and lungs[38]. In our study, there was a robust increase in the 3-OH-KYN/KYN ratio, pointing to the activation of KMO. In the presence of KMO, neurotoxic and excitatory products of KYN are synthesized, including 3-OH-KYN and QA that have been associated with free radical production and excitotoxicity, respectively[14]. In comparison, we observed a less pronounced increase in KYNA synthesis. These results suggest that CKD induced an imbalance between the neuroprotective KYNA and the above-mentioned neurotoxic products[31]. In fact, KYN and QA are not only neurotoxic but they have also been associated with anxiety in preclinical and clinical studies[44, 56], as opposed to the anxiolytic role of KYNA[43]. The predominance of anxiogenic mediators is further amplified by 5-HT deficiency due to the shift of Trp metabolism to the KYN pathway in inflammation, which further increases the sensitivity to anxiety[38].
Since HPA axis dysregulation has previously been implicated in CKD, we investigated on one hand if indeed the endocrine HPA axis activity is altered and on the other hand if the expression of CRH and its receptors is influenced by our CKD model.
As an indicator of the endocrine HPA axis, we measured the plasma corticosterone 8 weeks after CKD induction. Our results showed that in our model CKD did not alter plasma corticosterone concentration. These results are in line with the findings of Lu et al, in which plasma corticosterone has not changed in either male or female rats 7 weeks after 5/6 nephrectomy[49]. Another study has investigated plasma ACTH and corticosterone levels at multiple time points following 5/6 nephrectomy in rats, and the increase in ACTH peaked at 60 days post-operation, whereas the rise in corticosterone has only become significant 90 days after the nephrectomy. Based on these data, it is possible that our sample collection happened too soon for a significant elevation in plasma corticosterone to develop.
CRH is expressed in a variety of anxiety-related extrahypothalamic sites of the central nervous system, including the central amygdala (CeA)[40]. In our gene expression analysis, significant upregulations of Crh, Crhr1, and Crhr2 were observed in the amygdala in the CKD group, which is in accordance with the development of anxiety-like behavior. In a study by Cipriano et al., the intra-amygdalar administration of CRH resulted in a reduction in open arm time and entries in the EPM test, while a selective CRHR1 antagonist induced an anxiety-like effect, pointing to the involvement of amygdalar CRHR1 signaling in anxiety[16]. Moreover, the continuous overexpression of CRH by a lentiviral vector in the central nucleus of the amygdala has been associated with the dysregulation of the HPA axis, increased acoustic startle response, as well as depressive-like behavior[36]. Crhr2 upregulation can also be linked to anxiety, as urocortin 2 (a CRHR2 agonist) induced anxiety-like behavior, when injected into the medial amygdala in rats[1].
Contrary to the gene expression results, no significant differences were detected in the amygdalar protein expression of CRH and its receptors. Protein levels vary on a high dynamic range, influenced by the rate of translation and degradation, the former being the determining factor[68]. The rate of translation is regulated by a multitude of molecular mechanisms, one of which is transcription. Additionally, the activity of eukaryotic initiating factors (EIFs), structural features (e.g. internal ribosome-entry sequences, upstream open reading frames, and secondary or tertiary RNA structures), RNA-binding proteins, as well as microRNAs and small interfering RNAs are involved in the regulation of translation[24]. Any of these mechanisms might be responsible for our results, therefore it is possible that the increased expression of the Crh, Crhr1, and Crhr2 genes, that we detected in the amygdala, does not reflect in their protein concentration. However, it is also possible that the HPA axis dysregulation has not yet developed in our model, or it just started to develop at the time of sample collection, and later a significant change could have been detected. Nevertheless, further studies are needed to establish the role of the HPA axis in CKD-induced anxiety.