Epidemiologic data suggest that individuals at all stages of chronic kidney disease (CKD) have a higher risk of developing neuropsychiatric disorders, cognitive impairment, and dementia. This risk is generally explained by the high prevalence of both symptomatic and subclinical ischemic cerebrovascular lesions. However, other potential mechanisms, including cytokine/chemokine release, production of reactive oxygen species (ROS), circulating and local formation of trophic factors and of renin–angiotensin system (RAS) molecules, could also be involved, especially in the absence of obvious cerebrovascular disease. In this review, we discuss experimental and clinical evidence for the role of these mechanisms in kidney–brain cross-talk. In addition, we hypothesize potential pathways for the interactions between kidney and brain and their pathophysiological role in neuropsychiatric and cognitive changes found in patients with CKD. Understanding the pathophysiologic interactions between renal impairment and brain function is important in order to minimize the risk for future cognitive impairment and to develop new strategies for innovative pharmacological treatment.
Patients with renal diseases have a higher frequency of neuropsychiatric disorders and vice versa. There is evidence for a pathophysiological role of inflammation in both neuropsychiatric disorders and CKDs.
The cross-talk between the kidney and brain may include cytokine/chemokine release, production of ROS, circulating and local formation of trophic factors and of RAS molecules.
Understanding the interactions between the kidney and brain may lead to new strategies for innovative pharmacological treatment.
The vital roles of the kidney include filtration of the blood to maintain fluid and electrolyte balance and to remove blood waste, the stimulation of red blood cell production by hormones, the activation of vitamin D, the maintenance of bone health, and the releasing of hormones that mainly act in blood pressure control . The long-term loss of renal function and/or structure is a common condition characterized as chronic kidney disease (CKD) [2,3], which has an expressive high mortality . CKD is actually considered a global health priority not only because of the global increase in prevalence, but also due to the economic burden of end-stage renal disease (ESRD) [5–7]. CKD is one of the 20 commonest causes of death in the Global Burden of Disease Study 2010 [6,7]. In developed countries, the frequency of moderate and severe CKD in population-based studies reaches approximately 5–6% [8,9], being higher in low-income populations, in specific ethnic groups, and in older individuals [5,10,11].
As in every chronic condition, patients with CKD may suffer from limited functional capacity, impaired productivity, and reduced quality of life . Psychiatric disorders and cognitive impairment are highly prevalent in patients with CKD. Kimmel et al.  showed that patients with CKD had to be hospitalized for psychiatric disorders, particularly depression, dementia, and substance abuse, 1.5 to 3.0 times more than individuals with other chronic diseases. The prevalence of depression symptoms in patients with ESRD ranges between 20% and 25%, ranking them second only to hypertension in the roster of comorbidities observed in patients with ESRD [14,15]. Cognitive involvement and dementia are also commonly seen in patients with CKD, particularly at advanced stages . However, individuals with disease in any stage are susceptible to cognitive disorders associated with increased risk of death, lower levels of compliance to proposed treatment, higher incidence of cerebrovascular disease, and prolonged hospitalization [17,18]. Although these conditions are still underdiagnosed and undertreated in CKD patients .
The predisposing factors for alterations of central nervous system (CNS) functions in patients with CKD include sensory loss, cerebrovascular disease, subclinical metabolic disorders, and polypharmacy. Hospitalization and alterations in the metabolism of certain drugs (opioids, psychotropic drugs, antibiotics, and antiviral medication) and their toxic metabolites may also contribute [20,21]. The mechanisms involved in this process have not been completely elucidated but studies indicate that, in addition to neuronal damage induced by uremic toxins, the risk of cognitive involvement and dementia in these patients may be due to the high prevalence of symptomatic and asymptomatic brain ischemia . This vascular mechanism may explain the association between the risk factors affecting both the brain and the kidneys and its potential exacerbation in renal disease . However, other mechanisms including oxidative stress, immuno-inflammatory processes, trophic factors, and molecules related to the renin–angiotensin system (RAS) may also be involved in neurocognitive performance decline (Figure 1) [22–24]. Despite the fact that emerging studies have been supporting an immune-inflammatory-based process in CKD-associated cognitive and behavioral dysfunctions, it is worth noting that direct evidence linking this chronic condition to brain damage is still missing.
Immunological mechanisms of central nervous system and periphery communication
This review article will discuss experimental and clinical evidence for the role of these mechanisms in kidney–brain cross-talk and hypothesize potential pathways for the interactions between the kidney and brain and their pathophysiological role in neuropsychiatric and cognitive changes found in patients with CKD.
Immunological mechanisms of the central nervous system and periphery communication
The classical concept of ‘immune-privilege’ of the CNS has changed lately . The term privilege evokes a concept of advantage gained by an individual with respect to the common advantages of others. Therefore, the so-called ‘immune-privilege’ of the CNS has been considered an important factor in the protection against tissue injury during the course of inflammation, mostly for organs with poor capacity of regeneration. However, recent data have proven that this notion is far from absolutely true and that the ‘immune-privilege’ varies according to brain region and age of the individual . Indeed, the brain contains macrophages and dendritic cells, which are immune cells present in the choroid plexus and in the meninges. In addition, neurons and glial cells have receptors for pro-inflammatory cytokines and prostaglandins .
The cross-talk between the CNS and immune system has become increasingly evident in recent years . The immune-privilege of the CNS is severely undermined by inflammatory processes in the periphery of the body for several reasons: breakdown of the blood–brain barrier (BBB) reducing the immunosuppressive effects of the CNS microenvironment; local release of cytokines and chemokines; facilitation of antigen drainage to the periphery; the recruitment of dendritic cells; and the establishment of tertiary lymphoid tissue in the meninges . Brain function can also be modulated by diverse immune mediators produced at the periphery of the body including, for instance, pro-inflammatory cytokines synthesized by immune cells in the course of an infection . Cytokines can cross the BBB and reach the CNS leading to neuropsychiatric alterations that range from sickness behavior to anxiety or even psychosis .
The brain and the periphery interact in several ways (Figure 1). The first pathway is mediated by primary afferent nerves, which are activated by locally produced cytokines as, for instance, in a case of abdominal infection, with activation of the vagus nerve [29,30]. The second pathway is mediated by circulating pathogens that induce the production of pro-inflammatory cytokines by the macrophages residing in the circumventricular organs (CVO) and the plexus choroid . The absence of the BBB in CVO allows the propagation of cytokine signals through the brain . The third pathway involves the activation of interleukin (IL)-1 type 1 receptor, located on perivascular macrophages and endothelial cells of brain venules, by circulating cytokines resulting in the local production of prostaglandins [32,33]. Lastly, the fourth pathway includes the passage of pro-inflammatory cytokines from the systemic circulation to the brain through saturable transport (ST) systems in the BBB . These communication pathways between periphery and CNS result in the activation of microglial cells with the production of pro-inflammatory cytokines by means of a short- and a long-term mechanism. The short-term mechanism is elicited by afferent neural stimulus, whereas the long-term mechanism depends on the propagation of cytokine signaling within the brain [35,36].
Further evidence of the communication between the periphery and CNS is the so-called sickness behavior. Sickness behavior is defined by a set of behavioral changes, including neuropsychiatric symptoms, developed in sick individuals during the course of acute illness and/or inflammatory processes in the body . These changes have been associated with the action of pro-inflammatory cytokines in the brain, including IL-1β and tumor necrosis factor (TNF) . The transmission of peripheral responses to the CNS occurs via afferent nerves (neuronal route) located on the inflammatory site and also via the traditional endocrine route via blood (humoral route). Both routes transport peripherally synthetized cytokines to the brain. These cytokines stimulate CNS immune cells and activate the microglia, which in turn, increase local release of inflammatory molecules, prostaglandins, and other immune system products that ultimately affect brain functions [38,39]. These molecules act in the brain to induce common symptoms of sickness, including loss of appetite, sleepiness, withdrawl from normal social activities, fever, aching joints, fatigue, embezzlement of cognition, malaise, inability to concentrate, depression, and lethargy [38,39].
Sickness behavior is now recognized to be part of a motivational system that reorganizes the organism's priorities to facilitate recovery from an acute illness such as an infection . For instance, infectious agents display a variety of pathogen-associated molecular patterns (PAMPs) that are recognized by pattern recognition receptors (PRRs). These PRRs are expressed on both the surface by means of the Toll-like receptor (TLR)-4 and in the cytoplasm by nucleotide-binding oligomerization domain (Nod)-like receptors of cells of the innate immune system, primarily macrophages and dendritic cells. These cells initiate and propagate an inflammatory response by stimulating the synthesis and release of a variety of cytokines. Once an infection has occurred in the periphery, both cytokines and bacterial toxins deliver this information to the brain using both humoral and neuronal routes of communication. The binding of PRR activates the afferent vagus nerve, which communicates neuronal signals via the lower brain stem to higher brain centers including the hypothalamus and amygdala. Blood-borne cytokines initiate a cytokine response from vascular endothelial cells that form the BBB. Cytokines can also reach the brain directly by leakage of the BBB via CVO or by being synthesized within the brain. Although all cells within the brain are capable of initiating cytokine secretion, microglia have an early response to incoming neuronal and humoral stimuli. Inhibition of pro-inflammatory cytokines that are induced following bacterial infection block the appearance of sickness behaviors .
The fact that cytokines act in the brain to induce physiological adaptations that promote survival has led to the hypothesis that inappropriate, prolonged activation of the innate immune system may be involved in a number of neurologic diseases, ranging from Alzheimer's disease to stroke. Conversely, the newly defined role of cytokines in a wide variety of systemic comorbid conditions, including CKD, may begin to explain changes in the mental state of these subjects [12,17]. Indeed, the newest findings of cytokine actions in the brain offer some of the first clues about the pathophysiology of certain mental health disorders, including depression.
Potential mechanisms of kidney–brain cross-talk
It was well recognized that CKD directly influences the circulating levels of cytokines and chemokines, and indirectly activates immune cells from remote regions of the body. The cytokines most frequently associated with the pathogenesis of CKD are IL-1β, IL-6, TNF, and transforming growth factor-β (TGF-β) [40,41].
IL-1β is a member of the IL-1 family of cytokines, which comprises 11 molecules, including seven pro-inflammatory agonists (IL-1α, IL-1β, IL-18, IL-33, IL-36α, IL-36β, and IL-36) and four putative antagonists (IL-1 receptor antagonist (IL-1Ra), IL-36Ra, IL-37, and IL-38), with IL-1α and IL-1β the most prominent components [42,43]. IL-1β has a role in acute and chronic inflammation, binding to IL-1 receptor type 1 (IL-R1), and activating pro-inflammatory signaling pathways, mostly the nuclear factor-κB (NF-κB) [44,45]. This cytokine has been implicated in neuroinflammation and altered expression of IL-1β can be found in a number of neuropsychiatric disorders, including Alzheimer's disease , Parkinson's disease , and major depressive disorder [48,49].
IL-6 is an immune mediator with a pleiotropic effect on inflammation, immune response, and hematopoiesis. This cytokine is involved in B and CD8+ lymphocytes differentiation and inhibition of T-regulatory lymphocytes, but it also has a broad spectrum of actions in other tissues, ranging from hepatocytes and connective tissue  to cells of the nervous system. The influence of this cytokine in the CNS was established in the early 1990s . The studies point to a role in physiological processes of neuronal and glial differentiation, survival and production of trophic factors, as well as inflammatory processes within the CNS . In the latter context, there is plenty of data that implicate the inflammatory effect of IL-6 in the pathophysiology of neuropsychiatric disorders like Alzheimer's disease, Parkinson's disease, multiple sclerosis, major depressive disorder, and schizophrenia .
TNF is a major pro-inflammatory cytokine. Its production mainly comes from immune cells like macrophages, T and B lymphocytes, mast cells, and neutrophils, but there are many other cells that can produce this cytokine, including endothelial cells, smooth muscle cells, fibroblast, cardiomyocytes, osteoblasts, osteoclasts, dendritic cells, adrenocortical cells, adipocytes, astrocytes, microglial cells, and glomerular mesangial cells [54–56]. Binding of TNF to the TNF receptors (TNFR) type 1 and type 2 triggers intracellular signaling pathways through NF-κB, mitogen activated protein kinase (MAPK), and caspases that can promote inflammation and cell proliferation, differentiation, and apoptosis . In the CNS, TNF takes part in several physiological processes like modulating ionotropic glutamate receptors trafficking, and thereby influencing the synaptic transmission and plasticity. In pathological contexts, elevated levels of TNF can induce pro-inflammatory signaling in glial cells by stimulating the release of other pro-inflammatory mediators and reactive oxygen and nitrogen species (ROS/RNS) . In addition, TNF may also have an inhibitory effect on glutamate transporters, thus contributing to excitotoxicity . These pathological features of TNF are common to many neuropsychiatric disorders like Alzheimer's disease, Parkinson's disease, Huntington's disease, and major depressive disorder [60–65].
TGF-β is synthesized primarily by platelets, macrophages, monocytes, lymphocytes fibroblasts, and epithelial cells. TGF-β has a role in a number of physiological and pathological processes, acting as an embryogenic molecule, an anti-inflammatory cytokine and, mainly, as a fibrogenic factor, implicated in tissue remodeling and scar tissue formation . This cytokine has been implicated in physiological processes in the CNS by modulating synapses and by exerting protective effects in the brain .
The capacity of pro-inflammatory cytokines (IL-1β, IL-6, and TNF) to reciprocally stimulate each other's production and to cooperate in inflammation supports the hypothesis that the inflammatory response triggered by CKD may influence remote organs such as, for instance the brain, reinforcing the idea of kidney–brain inflammatory cross-talk (Figure 1). For example, the canonical activation of NF-κB by TNF and IL-1β on astrocytes induces the secretion of hemolymphopoietic cytokines, like IL-6 and IL-8, and colony-stimulating factors .
Another class of immune system-related molecules is the so-called neurotrophic factors. This class of molecules includes mainly the brain-derived neurotrophic factor (BDNF), the nerve growth factor (NGF), and the glial cell line-derived neurotrophic factor (GDNF). The neurotrophic factors enhance nerve cells’ growth potentialities through the interaction with two classes of receptors: the high-affinity protein tyrosine kinase receptors (TrkA, TrkB, and TrkC), which preferentially bind to specific neurotrophic factors; and the low-affinity pan-neurotrophin receptor p75NTR, which is associated with pro-apoptotic signaling .
The pathophysiology of neuropsychiatric disorders has been frequently associated with mechanisms related to decreased availability of neurotrophic factors for the affected neurons . For instance, a significant reduction of neurotrophic factor levels in the serum and cerebrospinal fluid has been reported in several conditions, including Alzheimer's disease, Parkinson's disease, Huntington's disease, and major depression [71–81].
Interestingly, the role of neurotrophic factors was also evaluated in the immune system. Indeed, neurotrophic factors and their receptors are expressed in lymphoid organs, in T and B lymphocytes, and in monocytes [82–84]. In addition, there is evidence for interactions between inflammatory molecules and neurotrophic factors. Mondelli et al.  showed that leukocytes obtained from patients at the first episode of psychosis have increased gene expression of pro-inflammatory cytokines, IL-6, and TNF, in line with reduced expression of BDNF. It can be hypothesized that cytokines can cross the BBB to interact with neurotrophic factors in the CNS [86–88], mostly considering that the BBB may be disrupted in pathological and inflammatory conditions (Figure 2) [89–91].
Potential mechanisms of kidney-brain cross-talk
Very few studies used experimental models of CKD to evaluate alterations in the CNS. However, there is some evidence supporting that CKD per se might favor the access of inflammatory molecules to the CNS. In this regard, Mazumder et al.  showed that Adult Swiss albino female mice with CKD induced by a high adenine diet exhibit a disruption of the BBB, ultimately facilitating the entrance of inflammatory molecules, like cytokines, in the CNS. Although, to date, there is no direct evidence supporting a role of neurotrophic factors in kidney–brain interactions. On the other hand, it is reasonable to hypothesize that, like cytokines, neurotrophic factors may play a role in CKD-associated neuropsychiatric disorders.
Oxidative stress is the imbalance between the production of free radicals and the removal of these molecules by antioxidant systems leading to damage to proteins, lipids, and nucleic acids with potential loss of function and apoptosis . Additionally, there is a close link between inflammation and oxidative stress that can be observed by mutual up-regulation between the two processes. The inflammatory response induces production of ROS and RNS as a result of cellular metabolism and the activation of innate immunity components . Also, free radicals function as signaling molecules to promote inflammation, as, for instance, the activation of redox-sensitive transcription factors like NF-κB promotes pro-inflammatory cytokine and chemokine production, leading to recruitment and activation of leukocytes and resident cells . In this context, studies have shown that oxidative stress is related with the onset and development of diverse diseases like CKD and neurodegenerative disorders (Figure 1) [95,96].
Animal models of CKD have been associated with altered activation of the nuclear factor-erythroid-2-related factor 2 (Nfr2) pathway. Nfr2 is one of the main transcription factors that regulate the expression of related genes of antioxidant systems [97,98]. Kim et al.  reported that male rats submitted to 5/6 nephrectomy present a progressive reduction in the Nfr2 nuclear content, accompanied by down-regulation of Nfr2 target gene products, including antioxidant enzymes such as catalase (CAT), superoxide dismutase (SOD), and glutathione peroxidase (GPX). Furthermore, these animals have activation of NF-κB, intense infiltration of mononuclear leukocytes, and up-regulation of monocyte chemoattractant protein-1 (MCP-1) and cyclooxigenase 2 (COX-2) mRNA, indicating an increased inflammatory response . Other studies support a role for uremic toxins in the genesis of oxidative stress in CKD . In this regard, in a mice model of subtotal nephrectomy, urea and uremic toxins stimulated mitochondrial ROS production .
In humans, measurements of antioxidant molecules support a role for oxidative stress in CKD. For instance, in a cohort study, Oberg et al.  reported that patients with CKD stages 3 to 5 have a pro-oxidant profile in comparison with healthy individuals, as revealed by the reduction in plasma protein-reduced thiol content and the increase in plasma levels of protein-associated carbonyl and of free F2-isoprostane content. Interestingly, patients undergoing hemodialysis presented muscle atrophy, which was associated with oxidative stress as indicated by a significant reduction in malondialdehyde content and CAT activity in line with an increase in total muscle glutathione and heat shock protein content. However, direct evidence for the role of oxidants in muscle degeneration in this population is still missing . In regard to uremic toxins, D'Apolito et al.  showed that in vitro exposition of human aortic endothelial cells to 20 nM of urea for 48 h increases mitochondrial ROS production with consequent activation of pro-inflammatory pathways .
RAS is classically conceived as a hormonal system responsible for blood pressure control and hydroelectrolyte balance . However, over the last decades, the comprehension of RAS has changed by the discovery of local RAS components in diverse organs and tissues, of other biologically active angiotensin peptides besides angiotensin II (Ang II), of new functions for angiotensin peptides, and of novel enzymes and angiotensin receptors (for review, see [105,106]). Among these discoveries, it should be mentioned the local role of Ang II as a mediator of inflammation and fibrosis, acting via angiotensin type 1 (AT1) receptors and the characterization of the heptapetide angiotensin-(1–7) [Ang-(1–7)] as a biologically active RAS mediator that frequently oppose Ang II actions acting via a Mas receptor [107–109]. Ang-(1–7) is mainly formed by the action of the enzyme homolog to angiotensin-converting enzyme named ACE2 on Ang II [110,111]. RAS is now considered a system composed by two opposing axis: the classical one, including the ACE-Ang II-AT1 receptor and the alternative one, comprising the ACE2–Ang-((1–7)–Mas receptor.
The role of RAS peptides in the interaction between the kidney and brain is supported by studies showing that the treatment with ACE inhibitors and AT1 receptor blockers, besides exerting renoprotection, also have beneficial actions in neurodegenerative disorders , (for review, see ). In this regard, Abdalla et al.  showed that ACE inhibition exerts beneficial effects on signs of neurodegeneration of aged Tg2576 mice as a transgenic animal model of Alzheimer's disease. Whole genome microarray gene expression profiling and biochemical analyses showed that the ACE inhibitor captopril normalizes the excessive hippocampal ACE activity and retards the development of signs of neurodegeneration by 6 months . The neuroprotective actions of captopril were accompanied by reduced amyloidogenic processing of the amyloid precursor protein and decreased ROS in the hippocampus . These results obtained with experimental models are corroborated by clinical studies. For instance, O'Caoimh et al.  reported that patients with Alzheimer's receiving ACE inhibitors have a reduced rate of functional decline. ACE inhibitors exerted neuroprotective actions not only in experimental Alzheimer's disease but also in a rat model of Parkinsonism . It was verified that the treatment with captopril reduces oxidative stress and protects dopaminergic neurons in a 6-hydroxydopamine rat model of Parkinsonism . Similar results have been obtained with the administration of AT1 receptor blockers in Alzheimer's disease, Parkinson's disease, stroke, traumatic brain injury, and spinal cord injury (for review, see ). These findings support that ACE inhibitors and AT1 receptor blockers, which are widely used for renal and cardiovascular diseases, may also exert neuroprotection in diverse neurodegenerative disorders. The general idea is that the reduction of Ang II formation (ACE inhibitors) or signal transduction (AT1 receptor blockers) may decrease ROS production, cytokines, and chemokine release and the local activation of immune system cells (Figure 1) [24,112].
Considering that the ACE2–Ang-(1–7)–Mas receptor axis generally opposed the actions of the ACE–Ang II–AT1 axis, it seems reasonable that the administration of Ang-(1–7) and/or the activation of ACE2 or Mas receptor may also exert neuroprotection. Recent studies corroborate this hypothesis. Wang et al.  showed that mice with genetic deletion of ACE2 exhibit impaired cognitive function, probably due to enhanced oxidative stress and decreased levels of BDNF mRNA and protein in hippocampus. Accordingly, intracerebroventricular infusion of Ang-(1–7) ameliorates cognitive impairment and memory dysfunction in a mouse model of Alzheimer's disease . In addition, hypertensive transgenic (mRen2) 27 rats with overactivity of the RAS present anxiety- and depression-like behaviors that are reversed by the intracerebroventricular infusion of Ang-(1–7) . The pre-treatment of these animals with the Mas receptor antagonist, A-779, abolished the anxiolytic-like action produced by the treatment with the ACE inhibitor enalapril suggesting that, at least in part, endogenous Ang-(1–7) contributes to the CNS effects of ACE inhibition . On the other hand, very few studies have evaluated the ACE2–Ang-(1–7)–Mas receptor axis in human diseases. In regard to neurodegenerative disorders, there is only one study that measured components of RAS in the blood of Parkinson's disease patients . It was found that circulating levels of Ang I, Ang II, and Ang-(1–7) are reduced in patients with Parkinson's disease when compared with age- and sex-matched healthy controls . Furthermore, among patients with Parkinson's disease, lower circulating levels of angiotensins were associated with increased severity of depressive symptoms. Future studies need to clarify the mechanisms and brain areas mediating neuroprotective effects of ACE2–Ang-(1–7)–Mas receptor axis. It also remains to be investigated how medications used to treat renal and cardiovascular disease may affect CNS functions and behavioral responses.
The missing link: insights from kidney–brain interaction in neuropsychiatric disorders
The vascular hypothesis for CKD-associated neuropsychiatric disorders
Neuropsychiatric disorders including cognitive decline, depression, and anxiety are common outcomes of CKD, which often hamper patients quality of life and also result in longer hospitalization and higher risk for mortality [17,120,121]. Despite the significant socioeconomic burden and the efforts to better understand the kidney–brain axis cross-talk, to date, direct evidence linking CKD to brain damage is still missing .
Over the past decades, some hypotheses have been raised, especially associated with cerebrovascular disease and accumulated uremic toxins, which include guanidino compounds and parathyroid hormone (for review see [123,124]). The vascular theory is based on kidney and brain hemodynamic similarities, since both are low resistance end organs exposed to high-volume blood flow and, consequently, more susceptible to vascular damage . Accordingly, magnetic resonance imaging (MRI) studies showed high occurrence of silent brain infarction in patients with CKD, supporting the vascular hypothesis [126–129]. For instance, approximately 50% of CKD patients presented ischemic white matter lesions in the MRI compared with 10% of the general population . The lack of a direct association between known vascular risk factors like diabetes and hypertension with CKD-related cognitive deficits , the occurrence of neuropsychiatric comorbidities in pediatric patients with CKD, preceding the vascular damage , as well as inconsistent findings regarding antihypertensive drugs beneficial effects in cognition , may indicate that other mechanisms underlie the CKD-associated CNS dysfunctions.
Potential role of inflammatory mediators in kidney–brain cross-talk
Cerebrovascular injuries in CKD have also been associated with retention of uremic toxins along with electrolyte imbalance, which ultimately leads to neuropsychiatric diseases, especially cognitive deficits and dementia . A more recent study reported several uremic toxins that potentially mediate the interactions between kidney and brain, which in turn, may influence CNS homeostasis. Among others, uric acid, indoxyl sulfate, p-cresyl sulfate, IL-1β, IL-6, TNF, and parathyroid hormone are likely to have an impact on cognition under uremic conditions . Moreover, uremic neurotoxic effects seem to be also mediated by guanidino compounds, including creatinine, guanidine, guanidinosuccinic acid, and methylguanidine. Importantly, a high concentration of those guanidino compounds was found in key brain regions involved in cognition and behavior, such as the thalamus, the mammillary bodies, and the cerebral cortex, of CKD patients . Although the potential role of uremic toxins in mediating the kidney–brain cross-talk has become clearer over the past years, the mechanisms by which these compounds may directly or indirectly influence the CNS function remain to be fully addressed. Significant differences in methodological approaches hampered robust conclusions regarding the uremic toxins potential mechanisms underlying the cerebrorenal interactions. Moreover, a wide range of evidence comes from in vitro studies or experimental models without renal failure, limiting the translation of these findings to human subjects with CKD . Inflammatory responses have often been reported in CNS and kidney injuries. Considering that inflammatory mediators such as cytokines participate in the peripheral and central systems communication, it is quite reasonable to hypothesize that they may be involved in kidney–brain interactions. A growing body of studies has been supporting these hypotheses, especially following acute kidney ischemia (for review see ). However, evidence regarding the inflammatory mediator's contribution in kidney–brain interactions during CKD remains elusive. Accordingly, a single elegant study demonstrated, by employing a progressive CKD model in rats, that increased levels of pro-inflammatory cytokines like IL-1β, IL-6, and TNF were associated with DNA damage in brain cells during the late stages of CKD . Moreover, 4 months after a 5/6 renal mass ablation, nephrectomized rats presented aversive-related memory and attention impairment in parallel with a significant increase in NF-κB and TNF expression in the hippocampus and frontal cortex. A decrease in the levels of soluble KLOTHO in the frontal cortex, a humoral factor that regulates, among others, the activity of ion channels and growth factor receptors such as insulin/insulin-like growth factor-1 (IGF1) receptors on the cell surface, was also found . In previous studies conducted with CKD patients, higher serum levels of the chemokine MCP-1 were associated with silent cerebral infarctions  and a serum proteomic profile consisting of the immune mediators IL-10 and C-reactive protein was 93% accurate in detecting mild cognitive impairment in CKD patients . Altogether, these studies provided evidence that inflammatory molecules might also mediate kidney–brain cross-talk during chronic events. However, further studies are urgently needed to better clarify this issue.
Putative role of oxidative stress in kidney–brain interactions
Oxidative stress, likely inflammation, is also a common event involved in brain and kidney damage and might potentially mediate the cerebrorenal communication. Emerging evidence has been supporting the role of ROS and RNS in CDK-associated cognitive and behavioral alterations. Accordingly, antioxidant drugs have significantly prevented the neuropsychiatric outcomes of CKD [138,139]. For instance, Western blot and immunohistology analysis revealed a significant increase in nitrotyrosine, an ultimate reactive and cytotoxic product generated by the interaction of NO and ROS, in the cerebral cortex of rats 6 weeks after 5/6 nephrectomy. Administration of the potent antioxidant lazaroid suppressed oxidative stress, as evidenced by normalization of the plasma lipid peroxidation product malondialdehyde, and reduced the concentration of nitrotyrosine in the cerebral cortex . A more recent study supported the involvement of oxidative stress in spatial working memory impairment following CKD induced by a left nephrectomy and 2/3 electrocoagulation of the right renal cortex. The antioxidant Tempol prevented cognitive dysfunction through inhibition of oxidative DNA damage in the hippocampus even without improvement of renal function . Interestingly, within 4 weeks, 5/6 nephrectomized rats exhibited significantly decreased exploratory behavior and locomotion along with a decrease in hippocampal levels of BDNF and with an increase in the serum levels of asymmetric dimethyl arginine (ADMA), a uremic toxin that acts as an endogenous inhibitor of NO. ADMA infusion alone in rats, without previous renal injury, decreased serum levels of BDNF, reproduced both behaviors and also developed a more pronounced anxiety-like behavior when compared with 5/6 nephrectomized rats. Depressive symptoms in CKD patients were also correlated with a significant reduction in the serum levels of BDNF and an increase in the levels of ADMA . Taken together, these studies provided further evidence that ROS and RNS lead to biochemical modification of brain in CKD, and might explain, at least in part, the neuropsychiatric disorders associated with chronic renal failure.
Evidence of RAS involvement in kidney–brain communication
Accumulation of uremic toxins has been considered as a putative cause of brain oxidative stress. Although the exact mechanism of the generation of oxidative stress remains to be fully elucidated . It has been reported that brain production of Ang II might be a risk factor for oxidative stress leading to cognitive dysfunction [142–144]. It is worth mentioning that some uremic toxins, like indoxyl sulfate, might activate local production of RAS metabolites at the CNS [145,146]. This pathway opens the door for the hypothesis that the brain RAS might play a role in CKD-associated cognitive and behavior dysfunctions. Accordingly, the administration of captopril, an ACE inhibitor, attenuated oxidative stress, ROS–NO interaction, and tyrosine nitration production in the cerebral cortex of rats 6 weeks after 5/6 nephrectomy . Administration of Telmisartan, an AT1 receptor blocker, prevented spatial memory impairment by decreasing brain oxidative DNA damage and lipid peroxidation, reinforcing the hypothesis that brain RAS is activated in CKD and possibly contribute to the associated cognitive decline .
The kidney and the brain interact in a strong and complex manner, often leading to significant neuropsychiatric comorbidities for patients with CKD. In this scenario, the comprehension of the mechanisms underlying kidney–brain communication paving the way for the development of appropriate therapeutic interventions is a scientific goal of highest priority. The immune system roles in brain dysfunctions associated with acute renal disease have been widely investigated; however, whether inflammatory processes underlie CKD-associated cognitive and behavioral alterations remains to be fully explored. Indeed, direct evidence linking CKD to brain damage is still missing. Herein, we provided an overview of inflammatory mediators potentially involved in the kidney–brain cross-talk and its associated pathologies. Since cytokines, chemokines, oxidative stress, and RAS components are commonly involved in kidney and brain damage, it is quite reasonable to hypothesize that they might mediate the cerebrorenal interactions contributing to the occurrence of neuropsychiatric disorders. However, despite the emerging evidence, significant differences in the methodological approaches in both clinical and experimental studies have hampered strengthening of conclusions. Further studies that systematically address this issue are urgently needed.
A.S.M., T.M.C., T.M.S.L.S., and R.N.F. drafted and revised the manuscript. A.C.S.S. drafted, revised, and edited the manuscript. All authors have read and approved the final version of the manuscript.
This work was supported by the Fundação de Amparo à Pesquisa do Estado de Minas Gerais [FAPEMIG, grant number PPM-00555/15]; and Conselho Nacional de Pesquisa [grant number 470472/2014-6].
The authors declare that there are no competing interests associated with this manuscript.
angiotensin converting enzyme
angiotensin converting enzyme 2
asymmetric dimethyl arginine
- Ang II
angiotensin type 1
brain-derived neurotrophic factor
chronic kidney disease
central nervous system
end-stage renal disease
mitogen activated protein kinase
nuclear factor-erythroid-2-related factor 2
pattern recognition receptor
reactive oxygen species
transforming growth factor-β
tumor necrosis factor