Nephrol Dial Transplant (2017) 1–9

doi: 10.1093/ndt/gfx039

Full Review

Urea and chronic kidney disease: the comeback of the century?

(in uraemia research)

Raymond Vanholder1, Tessa Gryp2 and Griet Glorieux2

1
Nephrology Section, Department of Internal Medicine, Ghent University Hospital, Ghent, Belgium and 2Laboratory for Bacteriology Research,
Department of Clinical Chemistry, Microbiology and Immunology, Ghent University, Ghent, Belgium

Correspondence and offprint requests to: Raymond Vanholder; E-mail: Raymond.vanholder@ugent.be

||
ABSTRACT ||
||
Urea, a marker of uraemic retention in chronic kidney disease ||
||
(CKD) and of adequacy of intradialytic solute removal, has trad- ||
itionally been considered to be biologically inert. However, a ||
||
number of recent experimental data suggest that urea is toxic at ||
concentrations representative for CKD. First of all, at least five ||
||
studies indicate that urea itself induces molecular changes ||
related to insulin resistance, free radical production, apoptosis
||
||
and disruption of the protective intestinal barrier. Second, urea ||
is at the origin of the generation of cyanate, ammonia and car-
||
||
bamylated compounds, which as such all have been linked to ||
biological changes. Especially carbamylation has been held re-
||
||
sponsible for post-translational protein modifications that are ||
||
involved in atherogenesis and other functional changes. In ob-
servational clinical studies, these carbamylated compounds
were associated with cardiovascular and overall morbidity and
mortality. These findings shed new light on the validity of Kt/
Vurea as a marker of dialysis adequacy. Yet, also the views that
the kinetics of urea are not representative of the kinetics of sev-
eral other uraemic retention solutes, and that urea cannot be
held responsible for all complex metabolic and clinical changes
responsible for the uraemic syndrome, still remain valid. Future
efforts to improve the outcome of patients with CKD might be
directed at further improving removal of solutes implied in the
uraemic syndrome, including but not restricted to urea, also tak-
ing into account the impact of the intestine and (residual) renal
function on solute concentration.
Keywords: ammonia, carbamylation, cyanate, urea, uraemic
toxicity
INTRODUCTION
The biological role of urea in chronic kidney disease (CKD)
and in the uraemic syndrome remains a matter of debate [1, 2].
The classical way of thinking was for many years that urea, in
spite of its worldwide use for the calculation of Kt/Vurea as a
marker of dialysis adequacy, is a relatively inert molecule that
exerts no major toxicity. This hypothesis was nourished by a
number of acute animal experiments, mostly from the 19th cen-
tury, showing no nominal toxic effects for urea [3], the experi-
ence of Johnson et al., who added urea to dialysate at
concentrations in excess of pre-dialysis values without a consist-
ent relationship between uraemic symptoms and urea exposure
|| [4], and two randomized controlled trials (the HEMO and
|| ADEMEX studies) finding no improved hard outcomes in spite
||
|| of increases in urea removal [5, 6].
|| In contrast, a randomized controlled study by Chertow et al.
||
|| showed a higher Kt/Vurea and improved outcomes in patients
|| treated by daily dialysis when compared with those on a stand-
||
|| ard thrice weekly dialysis regime [7], a finding that has been
||
|| cited as a possible indirect clinical argument in favour of the
|| toxicity of urea [3, 8]. However, these positive results are based
||
|| on a composite of hard and surrogate endpoints, and for daily
|| dialysis may be related to other factors than solute removal such
||
|| as less hypotensive episodes due to lower body weight gain in
|| between dialysis sessions, or to better removal of solutes other
||
|| than urea with potential impact on outcomes [9, 10].
||
|| More compelling than these clinical data are a number of re-
| cent experimental studies that have suggested that urea, either

C The Author 2017. Published by Oxford University Press

V 1
on behalf of ERA-EDTA. All rights reserved.
by itself or via post-translational protein modifications, induces ||
biochemical alterations with a potential impact on outcomes. ||
Of note, whereas most animal or clinical studies mentioned
||
||
above are either acute experiments or correlate instantaneous ||
urea concentrations to long-term outcomes, all effects that have
||
||
been found in these novel studies are very likely chronic and cu- ||
||
mulative, so that the negative or neutral outcomes mentioned ||
do not automatically support the idea that urea is inert. In this ||
||
publication, we summarize the most recent studies showing ||
that urea has a biological impact, and subsequently we will try ||
||
to define the exact meaning of urea and its removal in the con- ||
text of uraemia, taking into account this novel information. ||
||
||
||
||
CHARACTERISTICS OF UREA ||
||
||
Urea is a small water soluble molecule with a molecular weight ||
of 60 g/mol. It contains two nitrogen atoms and is the end- ||
||
product of protein and nitrogen metabolism. Urea is the com- ||
pound with the highest concentration found in the blood of ||
uraemic patients [11]. The serum concentration of urea is ex- |||
||
pressed either as a molar or as a mass concentration. Serum ||
urea mass concentration is either specified for the complete ||
||
urea molecule or for nitrogen equivalents [blood urea nitrogen ||
(BUN)] at a ratio of 60/28. The conversion between different
||
||
units is: BUN (mg/dL) ¼ 0.47  [urea in mg/dL] ¼ 2.8  [urea
in mmol/L]. Under normal conditions, BUN ranges from 6.1 to
20.2 mg/dL, which corresponds to urea concentrations of 13–
43 mg/dL or 2.2–7.2 mmol/L. BUN levels are markedly
increased in CKD patients, reaching in patients with end-stage
renal disease, pre-dialysis concentrations that can reach 10
times or more the upper limit of the normal range.
DIRECT TOXICITY OF UREA
At the end of previous century, several studies indicated that in
renal medullary collecting duct cells, urea instigates a chain of
molecular events, such as the induction of early growth re-
sponse protein-1 (egr-1) and immediate-early gene (IEG) tran-
scription factor [12, 13] (Table 1). The urea concentrations used
in those studies (200 mM) were substantially higher than those
observed in CKD, as these studies mimicked the osmolar condi-
tions of kidney medulla, where a high urea gradient is present.
Egr-1 subsequently has been linked to the activation of protein
kinase C (PKC) and of inositol triphosphate [20], two entities
that activate and regulate several biological mechanisms, among
which the inflammatory response [21].
Zhang et al. [14] subsequently demonstrated that urea treat-
ment of renal medullary cells at concentrations of 100–300 mM
upregulated the expression of growth-arrest and DNA-damage
inducible gene (Gadd153/CHOP), a transcription factor that is
responsive to oxidative stress. This effect was antioxidant sensi-
tive [14].
A study, sometimes referred to in support of the toxicity of
urea [15, 18], demonstrated that in atherosclerosis-prone apoli-
poprotein E knock-out mice the group with elevated urea due to
experimentally induced CKD showed more prominent
2 R. Vanholder et al.
atherosclerotic lesions than control animals [22]. This effect was
attenuated by the anti-oxidant N-acetylcysteine (NAC) [22].
However, in this study, urea was used as a marker of CKD sever-
ity and a tool to select animals for further study of the protective
effect of NAC and thus supposedly was not the only uraemic re-
tention solute to be elevated in the CKD group. Hence, based on
that study, there are in our opinion no reasons to accept a pro-
ven causal link between elevation of urea and the observed vas-
cular lesions. Accordingly, no correlation was found between
atherosclerotic plaque area and urea concentration [22].
Studies on the biological impact of urea at concentrations
relevant for CKD appeared, to the best of our knowledge, only
since 2010.
The first major breakthrough describing a direct biochemical
effect of urea was published by D’Apolito et al. [15]. These au-
thors demonstrated that when 3T3-L1 adipocytes were submit-
ted to disease-specific concentrations of urea, this resulted in
radical oxygen species (ROS) production, increased expression
of the insulin resistance-related adipokines, retinol binding pro-
tein 4 (RBP4) and resistin and increased O-linked b-N-acetyl-
glucosamine (O-GlcNAc) [15]. The latter effect generates a
decrease of insulin signal transduction and thus creates a state
of insulin resistance [15]. These in vitro changes were matched
by those observed in uraemic mice and normal mice infused
with urea [15], while this effect was attenuated by the anti-
|| oxidant, superoxide dismutase/catalase mimetic [15].
|| Trecherel et al. [16] studied the in vitro expression of pro-
||
|| apoptotic proteins in human smooth muscle cells, exposed to
|| urea or phosphate. Of both uraemic toxins, only urea caused
||
|| relevant modifications, which occurred in one of the proteins of
||
|| the B cell lymphoma-2 (BCL2) family, namely BCL-2-
|| associated death (BAD) promoter, a finding that conforms to
||
|| alterations observed by the same authors in uraemic mice [16],
|| and pointing to the potential of urea to induce apoptosis and
||
|| cell death in smooth muscle cells, a step at play in atherogenesis
|| and progression of atherosclerosis. Of note, BAD has been
||
|| linked as well to alterations in carbohydrate metabolism [23].
|| Vaziri et al. [17] tested the impact of urea on the integrity of
||
|| the intestinal epithelial barrier. Previous studies by the same au-
|| thors in CKD or in the presence of uraemic plasma had demon-
||
|| strated a disruption of intestinal barrier functions, potentially
|| impairing the protection against leakage of intestinal content
||
|| such as pro-inflammatory endotoxin into the body [24, 25]. Of
|| note, it has been demonstrated that CKD is linked to increased
||
|| endotoxinaemia even at the pre-dialysis stage [26, 27]. At the
|| molecular level, this derangement was attributed to a decrease
||
|| in expression of tight junction proteins [24, 25].
|| Although uraemic intestinal urea concentration has rarely
||
|| been studied, in an experiment in dogs it was shown to be as
|| high as in serum [28]. Hence, studies applying CKD concentra-
||
|| tions are also appropriate for conditions in the intestinal lumen.
|| The study by Vaziri et al. [17] tested in vitro the electrical resist-
||
|| ance of a monolayer of intestinal epithelium as a surrogate of in-
||
|| testinal barrier function in the absence or presence of urea at
|| uraemic concentrations, and urea appeared responsible for a
||
|| dose-dependent increase in epithelial leakiness. In addition, the
|| expression of the tight junction proteins occludin, claudin-1
||
and ZO-1 was decreased [17]. Urease, expressed by intestinal
Table 1. Direct biochemical modifications induced by urea (in vitro and animal experiments)

Study [Ref] Cell type or animal Change Urea concentration

A. In vitro experiments
Cohen et al. [12, 13] Medullary collecting duct cells Induction egr-1 200 mMa
Induction IEG transcription factor
Zhang et al. [14] Renal medullary cells Induction Gadd153/CHOP 100–300 mMa
D’Apolito et al. [15] Adipocytes Production ROS 20 mM
Expression RBP4b
Expression resistinb
Increase O-GlcNAc
Decrease insulin signal transduction
Trecherel et al. [16] Vascular smooth muscle cells Induction BAD protein 20 mM
Apoptosis
Vaziri et al. [17] Intestinal epithelial cells Decrease TER 42–72 mg/dL (7–12.3 mM)
Decrease abundance occludinc
Decrease abundance claudin-1c
Decrease abundance ZO-1c
D’Apolito et al. [18] Aortic endothelial cells Induction mitochondrial ROS 20 mM
Increased PKC activity
Increase O-GlcNAc
Increased expression VCAM-1
Increased expression MCP-1
Intracellular accumulation AGEs
Inhibition GADPH
Inhibition of PGI2-synthase
Generation of ER stress
Koppe et al. [19] Pancreatic b-cells Decrease glucose-stimulated insulin secretion 10–100 mMa
Increase oxidative stress Effect from 20 mM on
Increase O-GlcNAcylation
Decrease glucose utilization
Decrease PFK-1 activity
B. Animal experiments
D’Apolito et al. [15] Mouse Insulin resistance 2.9 increase versus control
Increase resistinb
Increase RBP4b
Koppe et al. [19] Mouse Glucose-stimulated insulin secretion 16–18 mmol
Increase oxidative stress
Increase O-GlcNAcylation
egr-1, early growth response protein-1; IEG, immediate-early gene; Gadd153/CHOP, growth-arrest and DNA-damage inducible gene; ROS, reactive oxygen species; RBP4, retinol bind-
ing protein 4; O-GlcNAc, O-linked b-N-acetylglucosamine; BAD, BCL-2-associated death promoter; TER, transepithelial electrical barrier; ZO-1, zonula occludens-1; PKC, protein kin-
ase C; VCAM-1, vascular cell adhesion molecule-1; MCP-1, monocyte chemotactic protein-1; AGEs, advanced glycation end products; GADPH, glyceraldehyde-3-phosphate
dehydrogenase; PGI2, prostacyclin; ER, endoplasmatic reticulum; O-GlcNAcylation, O-linked N-glucosamine acylation; PFK-1, phosphofructokinase-1.
a
Concentrations above uraemic range (underscore).
b
Markers of insulin resistance.
c
Tight junction proteins.

bacteria, which converts urea to ammonia, worsened this effect ||
[17] (for the role of ammonia, see below). ||
D’Apolito et al. subsequently assessed whether the same con-
||
||
centrations that had increased ROS production in adipocytes in ||
||
their previous study [15], also induced a similar effect in arterial ||
endothelial cells [18]. At a concentration of 20 mM, ROS gener- ||
||
ation was indeed induced, engendering further downstream ||
from the activation of other potentially pro-inflammatory and ||
||
pro-atherogenic pathways, such as increased activity of PKC
and of the hexosamine pathway as indicated by an increase of
O-GlcNAc, increased expression of the leukocyte adhesion mol-
ecules vascular cell adhesion molecule-1 (VCAM-1) and mono-
cyte chemotactic protein-1 (MCP-1), accumulation of advanced
glycation end products as well as inhibition of glyceraldehyde-
3-phosphate dehydrogenase (GADPH) [18]. In addition, the
study showed that urea also induced endoplasmatic reticulum
stress and inhibition of the anti-atherogenic enzyme prostacyc-
lin synthase [18]. Neutralization of ROS production annihilated
these deleterious effects. Although an animal arm of the study
demonstrated similar changes in uraemic mice, no data on
the direct toxicity of urea administration to animals with nor-
mal kidney function was reported in this study [18], in con-
trast to the previous publication by the same group on urea
toxicity [15].
Koppe et al. [19] tested the direct impact of urea on pancre-
atic b-cells. If islets from normal subjects were exposed to urea,
|| and also when normal mice were treated orally for 3 weeks with
|| urea and their islets were subsequently isolated, glucose-
||
|| stimulated insulin secretion was decreased, a finding analogous
||
|| to what was observed in 5/6 nephrectomized animals [19]. This
|| effect was coupled to an increase in oxidative stress and in pro-
||
|| tein O-linked N-glucosamine acylation (O-GlcNAcylation), as
|| well as to a disturbance of glucose utilization and phosphofruc-
||
|| tokinase-1 activity, which together with the impaired insulin se-
||
|| cretion could be reversed by inhibiting O-GlcNAcylation [19].
| All these data together suggest a direct effect of urea on the

Urea and CKD 3

pancreas, resulting in deficient insulin secretion and impaired
glycolysis. ||
||
In summary, at least five studies show a direct biochemical ||
effect of urea at uraemic concentrations (Table 1) [15–19]. ||
||
Changes to a large extent are linked to insulin resistance, in- ||
flammation and vascular damage, which all relate to the early ||
||
and increased morbidity and mortality of CKD [29–32]. ||
Biochemical alterations in vitro were similar across different cell
||
||
lines, for example, endothelium, pancreatic cells and adipocytes. ||
||
In at least two studies, the in vitro findings were confirmed after ||
oral urea administration or urea infusion to animals with nor- ||
||
mal kidney function [15, 19], although these confirmatory ex- ||
periments were confined to a limited number of parameters. In ||
||
those animal studies it was not clear whether the effect that was ||
observed was related to the same direct cellular impact as shown ||
||
in the in vitro studies, or to an effect of urea appearing in the in- ||
testine and affecting the intestinal barrier as observed by Vaziri ||
||
et al. [17]. Whatever the mechanism, all these studies point to a ||
patho-physiological role of urea in uraemia. ||
||
||
||
||
UREA AS A SOURCE OF CYANATE
Cyanate is a free radical that is in equilibrium with urea.
Generally, it is accepted that 0.8% of the molar concentration of
urea is converted into cyanate. Probably due to the increased
availability of urea, cyanate levels are also elevated in CKD [33].
The best known and most extensively studied biological ac-
tivity of cyanate is related to its induction of carbamylation (see
below). Some studies, however, also attribute direct biological
action (toxicity) to cyanate. Incubation of coronary endothelial
cells in the presence of cyanate decreased nitric oxide synthase
expression, and increased thrombogenic tissue factor and plas-
minogen activator inhibitor-1 expression [34]. In mice, admin-
istration of cyanate diminished the vasorelaxing response to
acetylcholine of aortic rings [34]. All these changes may con-
tribute to the vascular damage of uraemia.
Cyanate also dose-dependently decreased glucose-sensitive
insulin secretion in pancreatic islets [19]. A similar effect was
observed with urea per se, but the pathways involved in the
cyanate-dependent changes were different from those induced
by urea, suggesting that both compounds act independently
from each other via different mechanisms [19].
UREA AS A SOURCE OF CARBAMYLATED
COMPOUNDS
Carbamylation has been recognized as a post-translational
modification of amino acid and protein modification leading to
a plethora of biochemical alterations (Table 2) (as reviewed in
[44–46]). As shown by experiments in mice, carbamylation can
occur in a host of organs, with more substantial accumulation
in CKD than in animals with normal kidney function [36].
Traditionally, urea has been considered the major source of
cyanate and hence of carbamylated compounds, such as e-car-
bamyllysine (homocitrulline). More recently however, Wang
et al. [47] reported myeloperoxidase (MPO)-catalysed oxidation
of thiocyanate as a novel mechanism to which they attributed
greater strength, especially at sites of inflammation and athero-
genesis. Thiocyanate has been reported to be abundant in smok-
ers and is also retained in CKD [47, 48]. Corroborating the
intrinsic role of this pathway, homocitrulline was markedly
increased in atherosclerotic aortas of human MPO-transgenic
mice when compared with wild-type animals [47]. In addition,
|| incubation of low-density lipoprotein (LDL) cholesterol with
|| MPO and thiocyanate resulted in cholesterol accumulation and
||
|| foam cell formation in macrophages and in proliferation of
|| smooth muscle cells [47].
||
|| Carbamylated compounds interfere with organ and body
|| functions through multiple mechanisms. Carbamylated proteins
||
|| activate mesangial cells into a profibrogenic prototype, with a po-
|| tential to play a role in the progression of kidney failure [35].
||
|| Within this context, the protein modifications related to carba-
||
|| mylation affect long-lived extracellular matrix proteins such as
|| collagen [36, 37], a process mimicking in accelerated fashion the
||
|| phenomena of regular aging [37], and suggesting a long lasting
|| persistence in the body of carbamylated products. These struc-
||
|| tural collagen changes also imply a functional element.
|| Leukocytes respond in a different way to these carbamylated col-
||
|| lagens when compared with the genuine molecules [38].
|| Also, LDL is susceptible to carbamylation [39]. In vitro, car-
||
|| bamylated LDL dose dependently induced endothelial cell death
|| and smooth muscle cell proliferation [39], vascular cell changes
||
|| of relevance to atherogenesis. In uninephrectomized apolipo-
||
|| protein E deficient mice subjected to a high fat diet and oral ad-
| ministration of urea, carbamylated LDL increased about 8-fold

Table 2. Functional and structural changes induced by carbamylation

Study [Ref] Affected cells or molecule Modification

Shaykh et al. [35] Glomerular mesangial cells Activation into a profibrotic phenotype
Pietrement et al. [36] Collagen Long-lasting structural changes
Gorisse et al. [37] Modification leukocyte response
Jaisson et al. [38]
Ok et al. [39] Low-density lipoprotein Endothelial cell apoptosis
Apostolov et al. [40] Smooth muscle cell proliferation
Massy et al. [8] Increased atherosclerosis in atherosclerosis-prone mice
Molecular mechanisms inducing vascular damage
Sun et al. [41] High-density lipropotein Disturbance endothelial repair function
Stim et al. [42] Haemoglobin Not evaluated
Mun and Golper [43] Erythropoietin Disturbance of erythropoietic response

4 R. Vanholder et al.
compared with non-urea consuming animals [40]. After assess- || protein energy wasting by impeding amino acid incorporation
ment by either intravital echography or post-mortem macro- || into proteins [45], but vice versa, amino acid depletion could in
scopic investigation of the aortic wall after Sudan black staining || its turn diminish the protection of regular proteins against car-
||
for lipid deposits, the mice with high carbamylated LDL had || bamylation, by diminishing the possibility to deviate carbamy-
higher degrees of atherosclerosis [40]. Carbamylated LDL has
|| lation to free amino acids that are not incorporated into
||
equally been linked to a variety of molecular alterations leading || proteins.
to vascular damage such as an increase in endothelial adhesion
|| In a more extensive study on the same group of diabetic
||
molecules [8]. Also, high-density lipoprotein is carbamylated in || dialysis patients with follow-up of 4 years, carbamylated albu-
CKD and inhibits endothelial repair function [41].
|| min was correlated to cardiac stress markers troponin-T and N-
||
Haemoglobin is carbamylated in renal failure [42]. Exposure || terminal pro-B-type-natriuretic peptide (NT-proBNP) and was
||
of haemoglobin to urea resulted in a time-dependent modifica- || associated with a history of heart failure and arrhythmia [50]. In
tion, especially during the first 9 days [42]. Unfortunately, this || addition, it was strongly related to a number of hard cardiac
||
study did not assess whether these changes had a patho- || endpoints such as cardiovasular mortality, sudden cardiac death
physiological impact. || and risk of death from cardiac failure [50]. Statins were mainly
||
Erythropoietin can be carbamylated by in vitro exposure to || beneficial in the patient stratum with low carbamylated albumin
cyanate [43]. Compared with unmodified erythropoietin, sub- || [50].
||
cutaneous injection of carbamylated erythropoietin to Sprague– || In an assessment of a cohort of 347 haemodialysis patients,
Dawley rats showed an inadequate erythropoietic impact [43]. || mortality was highest in the tertile with the highest protein-
||
Altogether these experimental data suggest a profound biolo- || bound homocitrulline concentration [51]. Homocitrulline and
gical influence of protein carbamylation. The question arises in || urea concentration were correlated although the correlation was
||
how far these effects are corroborated by clinical data (Table 3). || weak, suggesting a role for causative factors other than urea,
In a clinical analysis that was part of the studies by Wang et || and/or the fact that homocitrulline is the reflection of long-term
||
al. [47] unraveling the thiocyanate-linked pathway of carbamy- || cumulated urea concentrations rather than of one single instant
lation, plasma protein-bound homocitrulline was an independ- || measurement. In addition, the correlation with urea was mark-
||
ent predictor for cardiovascular disease and death in subjects || edly stronger than that with any of the other markers of uraemic
undergoing cardiac catheterization. || retention considered [51].
||
In a group of 187 haemodialysis patients who were followed ||
for their outcomes, carbamylated albumin was correlated to ||
||
mortality within the first year after enrolment and also to time- || UREA AS A SOURCE OF AMMONIA
averaged BUN, suggesting a potential value of measurement of
||
||
concentration of carbamylated compounds as an indicator of || Urea is converted to ammonia by urease, which is expressed by
chronic urea exposure and accumulation [49]. This relationship
|| a number of intestinal bacteria [52]. Ammonia is in part con-
||
with mortality and BUN was validated and corroborated in an || verted to ammonium hydroxide, and both compounds lead to
independent population of 1161 German diabetic dialysis pa-
|| an increase of the pH in the intestinal lumen, mucosal irritation
||
tients [49]. The study also suggested a higher degree of albumin || and enterocolitis as additional and possibly more important
||
carbamylation with serum amino acid depletion, and the find- || causes of disruption of the intestinal epithelial barrier than the
ing of this clinical association was supported by studies in mice || effect of urea per se [17]. In vitro studies evaluating expression
||
on low protein diet, which next to amino acid deficiency, also || of intestinal epithelial tight junction proteins not only showed
showed higher degrees of albumin carbamylation [49]. Hence, || their depletion in the presence of urea, but also a further de-
||
it could be possible that amino acid carbamylation leads to crease when urease, inducing ammonia generation, was added

Table 3. Studies linking carbamylation to clinical outcomes cited in this review

Study [Ref] Tested carbamylated
compound
Wang et al. [47] Homocitrulline
Berg et al. [49] Carbamylated albumin
Drechsler et al. [50] Carbamylated albumin
Koeth et al. [51] Protein-bound
homocitrulline
NT-proBNP, N-terminal pro-B-type-natriuretic peptide; RCT, randomized controlled trial.
Studied population Associated parameters Study type
Subjects undergoing cardiac Cardiovascular events Case control
catheterization Death
Haemodialysis patients Mortality Observational cohort
Diabetic haemodialysis patients Averaged urea concentration
CKD patients
Diabetic haemodialysis patients Troponin-T Post hoc analysis RCT
NT-proBNP
History of cardiac arrhythmia
History of heart failure
Cardiovascular mortality
Sudden cardiac death
Death from congestive heart failure
Haemodialysis patients Mortality Observational cohort

Urea and CKD 5

[17]. AST-120, an oral sorbent that contains charcoal and cap- || Part of the biochemical impact of urea overlaps with that of
tures uraemic toxins and their precursors and also ammonia, || carbamylated compounds. Although it is difficult to discern be-
||
protects against this disruption of the intestinal barrier function || tween these two in vivo, the in vitro experiments clearly demon-
[53]. || strate the independent activity of both factors, which does not of
||
The production of ammonia and its deleterious effects is fur- || course exclude that both can also induce cumulative effects.
ther enhanced by a shift in the composition of intestinal micro- || CKD in these processes is at the origin of a vicious circle by
||
biota in favour of bacterial families containing urease [54]. The || inducing the progressive retention of urea, cyanate and thio-
role of urease in ammonia production is stressed in nephrec- || cyanate, which are directly toxic and/or the sources of carbamy-
||
tomized rats, in which intestinal ammonia was reduced by the || lated compounds, the activation of MPO, which catalyses
|| carbamylation, and changes in intestinal microbiota and func-
administration of a urease blocker, acetohydroxamic acid [55]. ||
|| tion, which further enhance inflammation (Figure 1). Some of
|| these processes are additionally promoted by causes of kidney
||
|| failure or its progression, including smoking, diabetes, obesity
DISCUSSION ||
|| and aging, altogether creating a morbid snowball effect. In this
The data reviewed in the present article compel us to reconsider || way there are several arguments favouring future strategies that
|| would result in better and more consistent removal of urea than
our current paradigm about the toxicity of urea. What has long ||
been considered as an inert molecule now seems to interfere with || what we attain today.
|| Yet, for the time being, and with the therapeutic arsenal we
a host of biochemical and organ functions, either directly or indir- ||
ectly via carbamylation and other mechanisms (Figure 1). One of
|| have available, it is unlikely that it will be possible in the near fu-
|| ture to decrease urea selectively. Almost all removal strategies for
the most striking aberrations caused by urea is insulin resistance, ||| end-stage kidney disease simultaneously affect a host of uraemic
a mechanism related to cardiovascular morbidity and mortality ||
|| solutes. It is even questionable whether it would be useful to aim
[31, 32]. If any specific approach were to be developed to select- || for a selective removal of specific uraemic toxins when consider-
ively decrease urea concentration, assessment of insulin resistance ||
|| ing the multiple retention compounds with biological potential
could be a first surrogate outcome to appreciate whether such || [56–59]. Thus, it seems more logical to enhance the removal of
intervention would be efficient, at the same time definitely con-
||
|| the broad array of uraemic toxins including, but not limited to,
firming in a clinical setting the toxicity of urea, which up to now || urea, by extracorporeal treatment, preserving kidney function or
||
has mainly been demonstrated experimentally. || by restraining intestinal generation, if such approaches are cost-
Other components that became apparent in playing a role in || effective [60].
||
the negative biological impact of urea are production of ROS || It becomes more and more evident that a substantial part of
and other inflammatory mechanisms. Like insulin resistance, ||
|| the biochemical impact of uraemia occurs via the intestine as an
also these elements are linked to cardiovascular damage and || intermediary. Although the metabolic generation of urea can be
mortality, which are all major causes of uraemic morbidity and
||
|| considered as quite straightforward because it simply is the end-
mortality [29, 30]. || product of amino acid metabolism, the gastrointestinal system
||
|| plays a crucial role by the digestion of proteins at the origin of
|| urea and the generation of ammonia, which in turn contributes
||
|| to the disintegration of the intestinal barrier. Optional interven-
|| tions allowing an impact on these processes are protein restric-
||
|| tion, combined or not to ketoanalogues, adsorption of intestinal
|| metabolites and modification of intestinal microbiota toward a
||
|| more favourable microbial content [61].
|| As many of the changes depicted in this review may be cu-
||
|| mulative and start occurring long before the dialysis stage, ef-
|| forts might be focused on preventing retention of uraemic
||
|| toxins including urea and/or their biological impact already at
|| the earlier stages of CKD [60].
||
|| It could be considered to decrease blood urea by creating an
|| intestinal shift in favour of urease-containing microbiota. A
||
|| number of in vivo and in vitro studies have confirmed this pos-
|| sibility [62–64]. However, urease transforms urea in ammonia
FIGURE 1: Pathways involved in the toxicity of urea. Left side (blue
||
|| (see above) and thus one toxin into another, which seems less
coloured track): cyanate/carbamylation pathway; middle (green col- || desirable.
ored track): direct toxicity of urea; and right (red coloured track): ||
|| The question should be raised again, in view of these novel
ammonia pathway. The general elements within these pathways are ||
indicated in black. Processes and conditions are in normal font, com- || data, whether Kt/Vurea is still a valid marker of dialysis ad-
pounds are in bold. Not included in the figure is a pathway of carba-
|| equacy. Although Kt/Vurea has always remained part of our clin-
||
mylation that is independent of urea, which involves transformation || ical dialysis practice since its development, if only because it
of thiocyanate (increased in smokers and CKD patients) into cyanate || may mimic removal of other vital small water soluble com-
||
by interference of hydrogen peroxide and myeloperoxidase [47]. pounds like potassium [1], with the novel data described in this

6 R. Vanholder et al.
review, the validity of Kt/Vurea becomes even more salient. Yet, || when end-stage renal disease patients were compared with nor-
the paradigm that there is more in uraemic retention and in sol- || mal controls [73].
ute removal by dialysis than what is represented by Kt/Vurea
||
||
also remains as valid as before, as this index does not necessarily ||
grasp the impact of other aspects related to uraemic solute con-
||
|| CONCLUSION
centration, such as dialyser pore size, solute kinetic behaviour, ||
|| In summary, an increasing number of experimental studies
adsorption, intestinal generation or metabolism [1]. Urea for ||
certain cannot be held responsible for all complex metabolic || point to the toxicity of urea. This effect is in part direct and in
||
and clinical changes constituting the uraemic syndrome [1], || part indirect via the generation of cyanate, carbamylated com-
which also includes a number of factors such as nutritional and || pounds and ammonia. Observational data suggest the impact of
||
volume status, electrolyte homeostasis, inflammation, blood || carbamylated compounds on clinical outcomes, but to the best
pressure or left ventricular hypertrophy. || of our knowledge, there are no controlled clinical studies cor-
||
It remains a matter of debate why, if urea is really that toxic, || roborating the toxic effects of urea per se.
the results of the HEMO and the ADEMEX studies did not ||
|| FUNDING
show a benefit when Kt/Vurea was increased [5, 6]. One of the ||
reasons could be that, since many of the effects summarized in || This work was supported by the Research Foundation
|| Flanders (FWO Vlaanderen) (grant No. G017815N to T.G.).
this review are long lasting, damage had become irreversible ||
once urea removal was increased at the dialysis stage, which in || T.G. is supported from this grant as a doctoral student.
||
addition occurred in prevalent patients, who on average had ||
been on dialysis already for 3.6 years [5]. Second, increasing Kt/ ||
Vurea should have had an impact not only on the removal of ||| CONFLICT OF INTEREST STATEMENT
||
urea, but of many other diffusible compounds as well, and some ||
of these substances may have been beneficial, so that increasing || None declared.
||
their elimination might have neutralized the removal of noxious ||
compounds such as urea. Third, specifically for the HEMO
||
||
study, the 30% increase of Kt/V urea had a much smaller effect || REFERENCES
on pre-dialysis concentrations of a variety of toxins, including
||
|| 1. Vanholder R, Glorieux G, Eloot S. Once upon a time in dialysis: the last days
urea, for which a decrease by only 9% was observed [65, 66]. || of Kt/V? Kidney Int 2015; 88: 460–465
Last but not least, also for the HEMO study, a substantial pro-
|| 2. Daugirdas JT. Kt/V (and especially its modifications) remains a useful meas-
||
portion of patients were on dialysis times of 3.5 h or less, even in || ure of hemodialysis dose. Kidney Int 2015; 88: 466–473
|| 3. Lau WL, Vaziri ND. Urea, a true uremic toxin: the empire strikes back. Clin
the group on high dialysis doses [5], whereas data from the || Sci 2017; 131: 3–12
Dialysis Outcomes and Practice Patterns Study showed that the || 4. Johnson WJ, Hagge WW, Wagoner RD et al. Effects of urea loading in pa-
||
observational association of Kt/Vurea with survival was to a large || tients with far-advanced renal failure. Mayo Clin Proc 1972; 47: 21–29
extent lost when dialysis time decreased to 3.5 h or less [67]. || 5. Eknoyan G, Beck GJ, Cheung AK et al. Effect of dialysis dose and membrane
|| flux in maintenance hemodialysis. N Engl J Med 2002; 347: 2010–2019
Urea is only one of the many small water soluble compounds ||
|| 6. Paniagua R, Amato D, Vonesh E et al. Effects of increased peritoneal clear-
[57]. Other solutes with the same properties are uric acid, oxalate, || ances on mortality rates in peritoneal dialysis: ADEMEX, a prospective,
potassium and several of the guanidines, including asymmetric || randomized, controlled trial. J Am Soc Nephrol 2002; 13: 1307–1320
dimethylarginine and symmetric dimethyl arginine, many of || 7. The FHN Trial Group; Chertow GM, Levin NW, Beck GJ et al. In-center
||
which have been linked to cardiovascular events. It is not really cer- || hemodialysis six times per week versus three times per week. N Engl J Med
tain whether the kinetics of urea during dialysis are representative || 2010; 363: 2287–2300
|| 8. Massy ZA, Pietrement C, Toure F. Reconsidering the lack of urea toxicity in
for those of the other molecules of the group. For the ones that || dialysis patients. Semin Dial 2016; 29: 333–337
have been assessed, such as the guanidines, it is not the case [68]. ||
|| 9. Barreto FC, Barreto DV, Liabeuf S et al. Serum indoxyl sulfate is associated
A recent addition to the group of small water soluble com- || with vascular disease and mortality in chronic kidney disease patients. Clin J
pounds is trimethylamine-N-oxide (TMAO), which was linked to || Am Soc Nephrol 2009; 4: 1551–1558
|| 10. Fagugli RM, De Smet R, Buoncristiani U et al. Behavior of non-protein-
atherogenesis for the first time only in 2011 [69] and recognized ||
as a uraemic retention product even more recently [70, 71]. || bound and protein-bound uremic solutes during daily hemodialysis. Am J
|| Kidney Dis 40: 339–347
Although TMAO generally is linked to mortality and cardiovas- || 11. Almeras C, Argiles A. The general picture of uremia. Semin Dial 2009; 22: 329–333
cular events [69, 71], some studies attribute the mortality linked || 12. Cohen DM, Gullans SR, Chin WW. Urea inducibility of egr-1 in murine
||
to TMAO to confounders like kidney dysfunction and poor || inner medullary collecting duct cells is mediated by the serum response
metabolic control [72]. The question of whether urea kinetics
|| element and adjacent Ets motifs. J Biol Chem 1996; 271: 12903–12908
|| 13. Cohen DM, Chin WW, Gullans SR. Hyperosmotic urea increases transcrip-
would be representative for those of TMAO remains partially un- ||
answered up till now. However, a study by Hai et al. [73] com-
|| tion and synthesis of Egr-1 in murine inner medullary collecting duct
|| (mIMCD3) cells. J Biol Chem 1994; 269: 25865–25870
pared normal renal and dialyser clearances of urea and TMAO. || 14. Zhang Z, Yang XY, Cohen DM. Urea-associated oxidative stress and
|| Gadd153/CHOP induction. Am J Physiol 1999; 276: F786–F793
Distribution volumes were also estimated, albeit only based on ||
pre- and post-dialysis concentrations. TMAO normal kidney || 15. D’Apolito M, Du X, Zong H et al. Urea-induced ROS generation causes insulin
|| resistance in mice with chronic renal failure. J Clin Invest 2010; 120: 203–213
clearance was substantially higher than that of urea, whereas its || 16. Trecherel E, Godin C, Louandre C et al. Upregulation of BAD, a pro-
dialysis clearance and distribution volume were lower, resulting || apoptotic protein of the BCL2 family, in vascular smooth muscle cells exposed
||
in a manifold larger proportional increase of TMAO than of urea to uremic conditions. Biochem Biophys Res Commun 2012; 417: 479–483

Urea and CKD 7

17. Vaziri ND, Yuan J, Norris K. Role of urea in intestinal barrier dysfunction 41. Sun JT, Yang K, Lu L et al. Increased carbamylation level of HDL in end-
and disruption of epithelial tight junction in chronic kidney disease. Am J
|| stage renal disease: carbamylated-HDL attenuated endothelial cell function.
||
Nephrol 2013; 37: 1–6 || Am J Physiol 2016; 310: F511–F517
18. D’Apolito M, Du X, Pisanelli D et al. Urea-induced ROS cause endothelial || 42. Stim J, Shaykh M, Anwar F et al. Factors determining hemoglobin carbamy-
dysfunction in chronic renal failure. Atherosclerosis 2015; 239: 393–400 || lation in renal failure. Kidney Int 1995; 48: 1605–1610
||
19. Koppe L, Nyam E, Vivot K et al. Urea impairs beta cell glycolysis and insulin || 43. Mun KC, Golper TA. Impaired biological activity of erythropoietin by cyan-
secretion in chronic kidney disease. J Clin Invest 2016; 126: 3598–3612 || ate carbamylation. Blood Purif 2000; 18: 13–17
20. Cohen DM, Gullans SR, Chin WW. Urea signaling in cultured murine inner || 44. Kraus LM, Kraus AP Jr. Carbamoylation of amino acids and proteins in ur-
medullary collecting duct (mIMCD3) cells involves protein kinase C, inosi-
|| emia. Kidney Int 2001; 59 (Suppl 78): S102–S107
||
tol 1,4,5-trisphosphate (IP3), and a putative receptor tyrosine kinase. J Clin || 45. Velasquez MT, Ramezani A, Raj DS. Urea and protein carbamylation in ESRD:
Invest 1996; 97: 1884–1889 || surrogate markers or partners in crime? Kidney Int 2015; 87: 1092–1094
21. Aljada A, Ghanim H, Mohanty P et al. Glucose intake induces an increase || 46. Kalim S, Karumanchi SA, Thadhani RI et al. Protein carbamylation in kid-
in activator protein 1 and early growth response 1 binding activities, in the
|| ney disease: pathogenesis and clinical implications. Am J Kidney Dis 2014;
||
expression of tissue factor and matrix metalloproteinase in mononuclear || 64: 793–803
cells, and in plasma tissue factor and matrix metalloproteinase concentra- || 47. Wang Z, Nicholls SJ, Rodriguez ER et al. Protein carbamylation links inflam-
tions. Am J Clin Nutr 2004; 80: 51–57 || mation, smoking, uremia and atherogenesis. Nat Med 2007; 13: 1176–1184
22. Ivanovski O, Szumilak D, Nguyen-Khoa T et al. The antioxidant N-acetyl-
|| 48. Koyama K, Yoshida A, Takeda A et al. Abnormal cyanide metabolism in ur-
||
cysteine prevents accelerated atherosclerosis in uremic apolipoprotein E || aemic patients. Nephrol Dial Transplant 1997; 12: 1622–1628
knockout mice. Kidney Int 2005; 67: 2288–2294 || 49. Berg AH, Drechsler C, Wenger J et al. Carbamylation of serum albumin as a
23. Danial NN, Walensky LD, Zhang CY et al. Dual role of proapoptotic BAD || risk factor for mortality in patients with kidney failure. Science Translat Med
in insulin secretion and beta cell survival. Nat Med 2008; 14: 144–153
|| 2013; 5: 175ra129
||
24. Vaziri ND, Goshtasbi N, Yuan J et al. Uremic plasma impairs barrier func- || 50. Drechsler C, Kalim S, Wenger JB et al. Protein carbamylation is associated
tion and depletes the tight junction protein constituents of intestinal epithe- || with heart failure and mortality in diabetic patients with end-stage renal dis-
lium. Am J Nephrol 2012; 36: 438–443 || ease. Kidney Int 2015; 87: 1201–1208
25. Vaziri ND, Yuan J, Rahimi A et al. Disintegration of colonic epithelial tight ||| 51. Koeth RA, Kalantar-Zadeh K, Wang Z et al. Protein carbamylation predicts
junction in uremia: a likely cause of CKD-associated inflammation. Nephrol
|| mortality in ESRD. J Am SocNephrol 2013; 24: 853–861
||
Dial Transplant 2012; 27: 2686–2693 || 52. Vaziri ND: Effect of synbiotic therapy on gut-derived uremic toxins and the
26. Goncalves S, Pecoits-Filho R, Perreto S et al. Associations between renal || intestinal microbiome in patients with CKD. Clin J Am Soc Nephrol 2016;
function, volume status and endotoxaemia in chronic kidney disease pa- || 11: 199–201
tients. Nephrol Dial Transplant 2006; 21: 2788–2794
|| 53. Vaziri ND, Yuan J, Khazaeli M et al. Oral activated charcoal adsorbent
||
27. McIntyre CW, Harrison LE, Eldehni MT et al. Circulating endotoxemia: a || (AST-120) ameliorates chronic kidney disease-induced intestinal epithelial
novel factor in systemic inflammation and cardiovascular disease in chronic || barrier disruption. Am J Nephrol 2013; 37: 518–525
kidney disease. Clin J Am Soc Nephrol 2011; 6: 133–141 || 54. Wong J, Piceno YM, Desantis TZ et al. Expansion of urease- and uricase-
28. Lee YT. Urea concentration in intestinal fluids in normal and uremic dogs. J
|| containing, indole- and p-cresol-forming and contraction of short-chain
||
Surg Oncol 1971; 3: 163–168 || fatty acid-producing intestinal microbiota in ESRD. Am J Nephrol 2014; 39:
29. Caravaca F, Martin MV, Barroso S et al. Do inflammatory markers add pre- || 230–237
dictive information of death beyond that provided by age and comorbidity in || 55. Swales JD, Tange JD, Evans DJ. Intestinal ammonia in uraemia: the effect of
chronic renal failure patients? Nephrol Dial Transplant 2006; 21: 1575–1581
|| a urease inhibitor, acetohydroxamic acid. Clin Sci 1972; 42: 105–112
||
30. Vanholder R, Massy Z, Argiles A et al..; European Uremic Toxin Work || 56. Vanholder R, Schepers E, Pletinck A et al. The uremic toxicity of indoxyl
Group. Chronic kidney disease as cause of cardiovascular morbidity and || sulfate and p-cresyl sulfate: a systematic review. J Am Soc Nephrol 2014; 25:
mortality. Nephrol Dial Transplant 2005; 20: 1048–1056 || 1897–1907
||
31. Yip J, Facchini FS, Reaven GM. Resistance to insulin-mediated glucose dis- || 57. Vanholder R, De Smet R, Glorieux G et al. Review on uremic toxins: classification,
posal as a predictor of cardiovascular disease. J Clin Endocrinol Metab 1998; || concentration, and interindividual variability. Kidney Int 2003; 63: 1934–1943
83: 2773–2776 || 58. Vanholder R, Baurmeister U, Brunet P et al.; European Uremic Toxin Work
32. Bodlaj G, Berg J, Pichler R et al. Prevalence, severity and predictors of
|| Group. A bench to bedside view of uremic toxins. J Am Soc Nephrol 2008; 19:
||
HOMA-estimated insulin resistance in diabetic and nondiabetic patients || 863–870
with end-stage renal disease. J Nephrol 2006; 19: 607–612 || 59. Duranton F, Cohen G, De Smet R et al.; European Uremic Toxin Work
33. Nilsson L, Lundquist P, Kagedal B et al. Plasma cyanate concentrations in || Group. Normal and pathologic concentrations of uremic toxins. J Am Soc
chronic renal failure. Clin Chem 1996; 42: 482–483
|| Nephrol 2012; 23: 1258–1270
||
34. El-Gamal D, Rao SP, Holzer M et al. The urea decomposition product cyan- || 60. Vanholder RC, Eloot S, Glorieux GL. Future avenues to decrease uremic
ate promotes endothelial dysfunction. Kidney Int 2014; 86: 923–931 || toxin concentration. Am J Kidney Dis 2015; 67: 664–676
35. Shaykh M, Pegoraro AA, Mo W et al. Carbamylated proteins activate glom- || 61. Vanholder R, Glorieux G. The intestine and the kidneys: a bad marriage can
erular mesangial cells and stimulate collagen deposition. J Lab Clin Med
|| be hazardous. Clin Kidney J 2015; 8: 168–179
||
1999; 133: 302–308 || 62. Mandal A, Das K, Roy S et al. In vivo assessment of bacteriotherapy on
36. Pietrement C, Gorisse L, Jaisson S et al. Chronic increase of urea leads to || acetaminophen-induced uremic rats. J Nephrol 2013; 26: 228–236
carbamylated proteins accumulation in tissues in a mouse model of CKD. || 63. Ranganathan N, Patel B, Ranganathan P et al. Probiotic amelioration of azotemia
PloS One 2013; 8: e82506
|| in 5/6th nephrectomized Sprague-Dawley rats. Scientif World 2005; 5: 652–660
||
37. Gorisse L, Pietrement C, Vuiblet V et al. Protein carbamylation is a hallmark || 64. Ranganathan N, Patel BG, Ranganathan P et al. In vitro and in vivo assess-
of aging. Proc Natl Acad Sci USA 2016; 113: 1191–1196 || ment of intraintestinal bacteriotherapy in chronic kidney disease. ASAIO J
38. Jaisson S, Lorimier S, Ricard-Blum S et al. Impact of carbamylation on type || 2006; 52: 70–79
||
I collagen conformational structure and its ability to activate human poly- || 65. Meyer TW, Sirich TL, Fong KD et al. Kt/Vurea and nonurea small solute
morphonuclear neutrophils. Chem Biol 2006; 13: 149–159 || levels in the hemodialysis study. J Am Soc Nephrol 2016; 27: 3469–3478
39. Ok E, Basnakian AG, Apostolov EO et al. Carbamylated low-density lipo- || 66. Meijers B, Vanholder R. HEMO Revisited: Why Kt/Vurea only tells part of
protein induces death of endothelial cells: a link to atherosclerosis in patients
|| the story. J Am Soc Nephrol 2016; 27: 3235–3237
||
with kidney disease. Kidney Int 2005; 68: 173–178 || 67. Saran R, Bragg-Gresham JL, Levin NW et al. Longer treatment time and
40. Apostolov EO, Ray D, Savenka AV et al. Chronic uremia stimulates LDL || slower ultrafiltration in hemodialysis: associations with reduced mortality in
carbamylation and atherosclerosis. J Am Soc Nephrol 2010; 21: 1852–1857 || the DOPPS. Kidney Int 2006; 69: 1222–1228
||
||
||

8 R. Vanholder et al.
68. Eloot S, Torremans A, De Smet R et al. Kinetic behavior of urea is different insufficiency and mortality risk in chronic kidney disease. Circ Res 2015;
from that of other water-soluble compounds: the case of the guanidino com-
|| 116: 448–455
||
pounds. Kidney Int 2005; 67: 1566–1575 || 72. Mueller DM, Allenspach M, Othman A et al. Plasma levels of
69. Wang Z, Klipfell E, Bennett BJ et al. Gut flora metabolism of phosphatidyl- || trimethylamine-N-oxide are confounded by impaired kidney function and
choline promotes cardiovascular disease. Nature 2011; 472: 57–63 || poor metabolic control. Atherosclerosis 2015; 243: 638–644
||
70. Kaysen GA, Johansen KL, Chertow GM et al. Associations of Trimethylamine || 73. Hai X, Landeras V, Dobre MA et al. Mechanism of prominent trimethyl-
N-Oxide with nutritional and inflammatory biomarkers and cardiovascular || amine oxide (TMAO) accumulation in hemodialysis patients. PloS One
outcomes in patients new to dialysis. J Ren Nutr 2015; 25: 351–356 || 2015; 10: e0143731
71. Tang WH, Wang Z, Kennedy DJ et al. Gut microbiota-dependent trimethyl-
||
||
amine N-oxide (TMAO) pathway contributes to both development of renal || Received: 4.2.2017; Editorial decision: 15.2.2017
||
||
||
||
||
||
||
||
||
||
||
||
||
||
||
||
||
|||
||
||
||
||
||
||
||
||
||
||
||
||
||
||
||
||
||
||
||
||
||
||
||
||
||
||
||
||
||
||
||
||
||
||
||
||
||
||
||
||
||
||
||
||
||
||
||
||
||
||
||
||
||
||
||
||