Profound Mishandling of Protein Glycation Degradation Products in Uremia and Dialysis
Stamatina Agalou*,
Naila Ahmed*,
Roya Babaei-Jadidi*,
Anne Dawnay and
Paul J. Thornalley*
* Department of Biological Sciences, University of Essex, Colchester, Essex; Renal Research Laboratory, St. Bartholomew’s and The Royal London School of Medicine and Dentistry, St. Bartholomew’s Hospital, London, United Kingdom
Address correspondence to: Dr. Paul J. Thornalley, Department of Biological Sciences, University of Essex, Wivenhoe Park, Colchester, Essex CO4 3SQ, UK. Phone/Fax: +44-1206-873010; thorp{at}essex.ac.uk
Received for publication August 4, 2004.
Accepted for publication February 16, 2005.
The aim of this study was to define the severe deficits of proteinglycation adduct clearance in chronic renal failure and eliminationin peritoneal dialysis (PD) and hemodialysis (HD) therapy usinga liquid chromatography-triple quadrupole mass spectrometricdetection method. Physiologic proteolysis of proteins damagedby glycation, oxidation, and nitration forms protein glycation,oxidation, and nitration free adducts that are released intoplasma for urinary excretion. Inefficient elimination of thesefree adducts in uremia may lead to their accumulation. Patientswith mild uremic chronic renal failure had plasma glycationfree adduct concentrations increased up to five-fold associatedwith a decline in renal clearance. In patients with ESRD, plasmaglycation free adducts were increased up to 18-fold on PD andup to 40-fold on HD. Glycation free adduct concentrations inperitoneal dialysate increased over 2- to 12-h dwell time, exceedingthe plasma levels markedly. Plasma glycation free adducts equilibratedrapidly with dialysate of HD patients, with both plasma anddialysate concentrations decreasing during a 4-h dialysis session.It is concluded that there are severe deficits of protein glycationfree adduct clearance in chronic renal failure and in ESRD onPD and HD therapy.
Both peritoneal dialysis (PD) and hemodialysis (HD) therapyas practiced currently have high mortality and morbidity—themedian survival time from commencement of dialysis therapy is5 to 8 yr, depending on other exacerbating factors (old ageand diabetes). Mortality and morbidity are associated with anincreased risk for cardiovascular disease. Inadequate removalof uremic toxins is a primary cause of uremia associated vasculardisease (1). Protein glycation adducts are a class of uremictoxin (2), but it is currently unclear how effective conventionalHD and PD therapies are in removing glycation adducts.
Glycation of proteins is a complex series of parallel and sequentialreactions collectively called the Maillard reaction. It occursin all tissues and body fluids. Early stage reactions with glucoselead to the formation of the early glycation adduct fructosyl-lysine(FL), and later stage reactions form advanced glycation endproducts (AGE) (3). FL degrades slowly to form AGE. Glyoxal(G), methylglyoxal (MG), and 3-deoxyglucosone (3-DG) are alsopotent glycating agents that are formed by the degradation ofglycated proteins, glycolytic intermediates, and lipid peroxidation.They react with proteins to form AGE directly (Figure 1, a andb). Important AGE quantitatively are hydroimidazolones derivedfrom arginine residues modified by glyoxal, MG and 3-DG, G-H1,MG-H1, and 3DG-H, respectively (Figure 1c). Other importantand widely studied AGE are N-carboxymethyl-lysine (CML), N-carboxyethyl-lysine(CEL), and pentosidine (4) (Figure 1, d and e). Proteins alsosuffer oxidative and nitrosative damage forming methionine sulfoxide(MetSO) (5) and 3-nitrotyrosine (3-NT) (6) (Figure 1f). Glycationadduct residues are formed by the physiologic glycation of endogenouscellular and extracellular proteins and are also present iningested food (7). Glycation free adducts are found in plasma,urine, and other physiologic fluids. They originate from theturnover of endogenous glycated proteins by cellular proteolysisand from food (4,8). Glycation reactions are increased in uremia,where there are increased concentrations of many -oxoaldehydes,particularly glyoxal, MG, and 3-DG (9,10). -Oxoaldehydes arealso present in PD fluids, formed by thermal sterilization (11),and introduced into the peritoneal cavity, they add to the increasedglycation potential in ESRD.
Figure 1. Protein glycation in physiologic systems. (a) Pathways for the formation of advanced glycation endproducts (AGE). (b) -Oxoaldehyde glycating agents. (c) Hydroimidazolone AGE: glyoxal-H1 (G-H1; 214 Da), methylglyoxal-H1 (MG-H1; 228 Da), and 3-deoxyglucosone-H (3DG-H; 318 Da). (d) Monolysyl glycation adducts: fructosyl-lysine (FL; 308 Da), N-carboxymethyl-lysine (CML; 204 Da), and N-carboxyethyl-lysine (CEL; 218 Da). (e) Cross-links and fluorophores: pentosidine (379 Da), (methylglyoxal-derived lysine dimer (MOLD; 341 Da), and argpyrimidine (254 Da). (f) Oxidation and nitration markers: methionine sulfoxide (MetSO; 165 Da) and 3-nitrotyrosine (3-NT; 226 Da). (c through f) Protein glycation, oxidation, and nitration residues are shown. For the corresponding free adducts at physiologic pH, the N-terminal amino group is protonated –NH3+ and the C-terminal carbonyl is a carboxylate –CO2– moiety.
The normal high renal clearance of AGE free adducts (4) is impairedmarkedly in ESRD; hence, the concentrations of free AGE in plasmaare expected to increase. In this study, we evaluated the characteristicsof glycation free adduct excretion of patients who had chronicrenal failure (CRF) with mild uremia and patients who had ESRDwith severe uremia and were on HD and PD therapies.
Clinical Sampling: Patient Characteristics and Blood and Urine Sample Collection
Venous blood samples were drawn from normal healthy subjects,untreated CRF patients, and ESRD patients into tubes with heparinanticoagulant. Subject characteristics are given in Table 1.Renal creatinine clearance for all dialysis patients was <10ml/min. Blood samples and aliquots of dialysate from HD patients(three with diabetes) were collected at the start and at theend of a 4-h dialysis session using a polysulfone membrane.Blood samples from PD patients (two with diabetes) were collected2 h after introduction of PD fluid into the peritoneal cavityduring a peritoneal equilibration test (PET) with peritonealdialysate collected after 0, 2, and 4 h of dwell time. Peritonealdialysate was also collected for approximately 12 h during thenight preceding the test and for the 24 h before this. Peritonealdialysate from four exchanges was pooled to estimate the 24-hexcretion flux. The PD fluids used by patients in this studywere single-compartment fluids (Dianeal; Baxter Healthcare Corporation,Deerfield, IL) that contained glucose osmolyte. Analyte estimatesfor PD and HD patients with diabetes were not outliers fromPD and HD patients without diabetes, respectively. Blood cellswere sedimented by centrifugation (2000 x g, 10 min), and theplasma was removed and immediately frozen at –80°C.Urine samples were collected at ambient temperature over 24h from normal healthy control subjects, CRF patients and PDpatients with residual diuresis (<10% of analyte amountswere lost during this period). Plasma and urine samples werestored at –80°C before analysis. The study was approvedby East London and The City Health Authority Research EthicsCommittee (London, UK), and written informed consent was givenby all participants and conformed to the Declaration of Helsinki.
Table 1 Clinical data for normal healthy control subjects and patients with CRF and ESRD on PD or HD therapya
Other Clinical Markers
Protein glycation adducts were assayed by liquid chromatographywith triple quadrupole mass spectrometric detection (LC-MS/MS)(4). The interbatch coefficients of variation were 10%, andanalyte estimates corroborated with independent measurementswhen these were available (6,12). Free glycation adducts weredetermined by assay of analytes in ultrafiltrate (12 kD filtercutoff, 50-µl aliquot) of plasma, urine, and dialysate.Glycation adduct residues of plasma protein were determinedin exhaustive enzymatic digests (50-µg protein equivalent)(4,13). Routine clinical service methods were used for urineand plasma creatinine (Jaffe rate method) and serum albumin(BCG colorimetric assay) on an Olympus analyzer. Plasma monocytechemotactic protein-1 (MCP-1) and TGF- were assayed in plateletfree plasma by double-antibody ELISA (R & D Systems, Abingdon,UK).
Statistical Analyses
Renal clearance of analytes (ml/min) for control subjects andCRF patients was determined as [Analyte]urine x urine volume/([Analyte]plasmaxurine collection time). Clearance of analytes in PD patientswas calculated similarly using both urine and peritoneal dialysateoutputs. Significance of difference between mean and medianAGE concentrations was determined using t test and the MannWhitney U test, respectively. Correlation analysis was performedby calculating Spearman statistic. Mass transfer area coefficients(MTAC) were deduced by the simplified Garred equation (14).
Detection of Glycation, Oxidation, and Nitration Free Adducts by LC-MS/MS in Plasma and Urine of Control Subjects and CRF Patients and Plasma, Urine, and Dialysate of ESRD Patients on HD or PD Therapy
Protein glycation, oxidation, and nitration free adducts weredetected by LC-MS/MS in plasma and urine of normal healthy controlsubjects, patients with CRF, and ESRD patients on PD and HDtherapy. They were also detected in peritoneal dialysate andhemodialysate of ESRD patients (Figure 2, a through h). Analyticalchromatograms of the hydroimidazolone MG-H1 showed the expectedpartially resolved pair of epimers (Figure 2, c and d), andthe hydroimidazolone 3DG-H showed the expected three resolvedstructural isomers (4) (Figure 2, e and f). The concentrationsof 3-NT were similar to those reported by others using massspectrometric techniques considered to be artifact-free (6).
Figure 2. Specimen analytical chromatograms in the determination of protein glycation adducts by liquid chromatography with triple quadrupole mass spectrometric detection in ESRD patients. CML (a) and [13C6]-CML (10 pmol; b) in plasma protein hydrolysate of a hemodialysis (HD) patient, Rt 7.9 min. MG-H1 (c) and [15N2]-MG-H1 (50 pmol; d) in peritoneal dialysis (PD) dialysate at 2 h of dwell time, Rt 24.0 (MG-HA) and 24.2 (MG-HB) min. 3DG-H (e) and [15N2]-3DG-H (50 pmol; f), isomers 1, 2, and 3 (as indicated) in hemodialysate, Rts 24.6, 25.1, and 23.8 min, respectively. Argpyrimidine (g) and [15N2]argpyrimidine (50 pmol; h) in HD patient plasma filtrate after a dialysis session, Rt 10.1 min. Chromatographic conditions were described in the Materials and Methods section.
Glycation, Oxidation, and Nitration Free Adducts in Plasma and Urine of Control Subjects and CRF Patients
The AGE free adducts MG-H1 and 3DG-H were found in the highestconcentration in blood plasma of control subjects, 122 nM, whichwas 133- and 46-fold higher than the concentrations of pentosidineand 3-NT free adduct (Table 2). Most glycation free adductswere increased in CRF patients, with respect to normal controlsubjects. The increases were two-fold for FL, four-fold forCML, four-fold for CEL, four-fold for G-H1, five-fold for MG-H1,two-fold for 3DG-H, and two-fold for argpyrimidine. This wasassociated with a marked decline in renal clearance of glycationadducts but similar 24-h excretion rate, with respect to controlsubjects (Table 3). The plasma concentration of glycation andoxidation free adducts correlated negatively with creatinineclearance for CML (r = –0.94, P < 0.01), MG-H1 (r =–0.79, P < 0.05), and 3DG-H (r = –0.75, P = 0.05),consistent with a decrease in GFR leading to the accumulationof glycation free adducts in plasma. Glycation free adductsare the major form of glycation adduct excretion in the urineof healthy subjects (4). The major glycation free adducts quantitativelyexcreted in the urine of control subjects and CRF patients wereFL, hydroimidazolones, CML, and CEL.
Table 2 Protein glycation, oxidation, and nitration free adduct concentrations in plasma of normal, healthy control subjects and patients with CRF and ESRD on PD or HD therapya
Table 3 Renal clearance of free glycation, oxidation, and nitration adducts of normal healthy control subjects and CRF patients, and urinary excretion rates of normal healthy control subjects, CRF patients, and ESRD patients on PD therapya
Glycation, Oxidation, and Nitration Free Adducts in Plasma, Urine, and Dialysate of PD Patients
In PD patients, the concentrations of most protein glycationfree adducts in blood plasma were increased, with respect tocontrol subjects. The increases were six-fold for CML, 10-foldfor CEL, four-fold for G-H1, 18-fold for MG-H1, eight-fold for3DG-H, five-fold for pentosidine, and 51% for 3-NT (Table 2).Protein glycation, oxidation, and nitration free adducts wereexcreted by transfer into the peritoneal cavity and eliminationin the peritoneal dialysate and by residual diuresis. Althoughthe clearance of most protein glycation, oxidation, and nitrationfree adducts in PD patients was decreased with respect to controlsubjects, the 24-h excretion rates were increased: G-H1 35%;FL, CML, CEL, argpyrimidine, and methylglyoxal-derived lysinedimer (MOLD) approximately two-fold; 3DG-H and pentosidine three-fold;3-NT four-fold; MG-H1 nine-fold; and MetSO 16-fold (Table 3).The proportion of protein glycation, oxidation, and nitrationfree adduct excreted in the peritoneal dialysate was 65% forFL, 76% for CML, 71% for CEL, 78% for G-H1, 78% for MG-H1, 75%for 3DG-H, 90% for argpyrimidine, 74% for MOLD, 46% for pentosidine,96% for MetSO, and 65% for 3-NT (two anuric patients were excludedfrom this calculation).
The concentration of most protein glycation, oxidation, andnitration free adducts in peritoneal dialysate increased withincreasing dwell time, except for MOLD and 3-NT, which did notincrease significantly (Figure 3 and Figure 4). From the 4-hPET, MTAC were deduced for 3DG-H and pentosidine free adductswhere peritoneal dialysate concentrations were consistent withequilibration with the plasma concentration. For these and otheranalytes, dialysate/plasma concentration (D/P) ratios were deduced(Table 4). At the 4- and 12-h dwell times, the concentrationsof glycation free adducts FL, CML, CEL, G-H1, MG-H1, and MetSOin the peritoneal dialysate exceeded greatly the concentrationsof the corresponding free adduct in blood plasma determinedat the 2-h dwell time point; hence, the D/P ratios were >1,and there were high peritoneal clearances (Table 4). The concentrationof 3DG-H free adduct in peritoneal dialysate at 12 h of dwelltime exceeded the concentration of 3DG-H free adduct in bloodplasma. These data suggest intraperitoneal formation of glycationand oxidation free adducts within the PD fluid dwell time.
Figure 3. Concentrations of protein glycation free adducts in peritoneal dialysate of PD patients showing the effect of PD fluid dwell time. (a) FL. (b) CML. (c) CEL. (d) G-H1. (e) MG-H1. (f) 3DG-H. (g) pentosidine. (h) MOLD. Data are mean ± SEM (n = 8). The dotted line is the corresponding free adduct concentration in plasma at the 2-h dwell time in the peritoneal equilibration test (PET).
Figure 4. Protein oxidative and nitrosative markers in plasma of normal healthy control subjects and chronic renal failure (CRF) patients and plasma and dialysate of PD patients. (A and B) Concentration of MetSO and 3-NT free adducts in peritoneal dialysate with dependence on PD fluid dwell time. The dotted line is the corresponding free adduct concentration in plasma at the 2-h dwell time in the PET. (C and D) Concentration of MetSO and 3-NT residues of plasma protein. CON, controls; PD (2), PD patients at 2 h in the PET; HD (0) and HD (4), HD patients before and after a 4-h dialysis session, respectively. Data are mean ± SEM (n = 7 to 8). *P < 0.05, **P < 0.01, and ***P < 0.001, with respect to control subjects.
Table 4 Dialysate/plasma concentration ratios and mass transfer area coefficients for the PETa
Glycation, Oxidation, and Nitration Free Adducts in Plasma and Dialysate of HD Patients
In HD patients, the plasma concentrations of glycation freeadducts before a dialysis session were increased markedly, withrespect to control subjects. The increases were three-fold forFL, 12-fold for CML, 21-fold for CEL, seven-fold for G-H1, 40-foldfor MG-H1, 10-fold for 3DG-H, four-fold for argpyrimidine, three-foldfor MOLD, and five-fold for pentosidine (Table 2). Similar highconcentrations of glycation free adducts were found in the HDdialysate at the start of the dialysis session, except for MG-H1and MOLD, reflecting rapid equilibration of glycation free adductsacross the dialysis membrane (Figure 5). Dialysis for 4 h normalizedthe plasma concentrations of pentosidine and MOLD free adductsand reversed the increases of other glycation free adducts inblood plasma by 54 to 93% but failed to normalize them (Table 2,Figure 5). The plasma concentrations of MetSO and 3-NT freeadducts in HD patients were increased but not significantly(Table 2).
Figure 5. Concentrations of protein glycation free adducts in plasma and hemodialysate of HD patients. (a) FL. (b) CML. (c) CEL. (d) G-H1. (e) MG-H1. (f) 3DG-H. (g) Pentosidine. (h) MOLD. Key: PCON, control plasma ultrafiltrate; PHD (0), plasma ultrafiltrate of HD patients at the start of dialysis; PHD (4), plasma ultrafiltrate of HD patients after 4 h of dialysis; DL (0), dialysate at the start of dialysis; DL (4), dialysate after a 4-h dialysis session. Data are mean ± SEM (n = 8). **P < 0.01 and ***P < 0.001, plasma concentrations with respect to control subjects. oP < 0.05, ooP < 0.01, and oooP < 0.001 for plasma and dialysate at the end of a dialysis session with respect to the corresponding concentration at the start of the dialysis session; +P < 0.05 and +++P < 0.001 for dialysate at the start of a dialysis session with respect to plasma at the start of dialysis.
The glycation free adduct reduction ratio (RR) was calculatedas the fall in concentration between the start and the end ofdialysis, expressed as a percentage of the concentration atthe start. Comparison was made with the RR for urea, which isan established measure of dialysis adequacy in a single sessionwith a minimum target of 65%. The urea RR was 69 ± 8%.Free adduct RR values were 84 ± 5% for CML and 77 ±9% for MOLD (greater than urea, P < 0.001 and P < 0.05);74 ± 8% for MG-H1, 74 ± 10% for 3DG-H, and 72± 9% for pentosidine (all not significant); and 39 ±19% for FL, 54 ± 17% for CEL, 52 ± 9% for G-H1(lower than urea, P < 0.01, P < 0.05, and P < 0.001).
Glycation, Oxidation, and Nitration Adduct Residues in Plasma Proteins of Normal Control Subjects, CRF Patients, and ESRD Patients on HD or PD Therapy
The plasma protein content of FL residues in control subjectswas 0.83 mmol/mol lys. This was decreased to 0.34 mmol/mol lys(–59%) in CRF patients and to 0.39 mmol/mol lys (–53%)in HD patients but was not decreased significantly in PD patients(Figure 6a). FL residues degrade to form CML residues (15).The plasma protein content of CML residues was 0.039 mmol/mollys in control subjects and increased two-fold in CRF and PDpatients and three-fold in HD patients. The plasma protein CMLresidue content of HD patients did not change during a dialysissession, but it was significantly higher than in CRF and PDpatients (Figure 6b). There was a lower plasma protein contentof CEL residues than CML residues in control subjects (0.021mmol/mol lys), and this was increased two-fold in all uremicpatients (Figure 6c). The plasma protein content of G-H1 residueswas 0.049 mmol/mol arg in control subjects and was increasedsignificantly in HD patients only after a dialysis session (51%;Figure 6d). There was a high plasma protein content of MG-H1residues in control subjects: 0.67 ± 0.13 mmol/mol arg.This was not increased in CRF patients but was increased approximatelytwo-fold in PD and HD patients. The plasma protein content ofMG-H1 residues was also decreased to control levels during adialysis session in HD patients (Figure 6e). The plasma proteincontent of 3DG-H residues in control subjects was lower thanMG-H1 residues—0.37 ± 0.04 mmol/mol arg—andwas increased two-fold in CRF and PD patients and three-foldin HD patients before dialysis (Figure 6f). The protein cross-links,pentosidine, and MOLD were minor glycation adduct residues quantitativelyin plasma protein of control subjects (0.0104 ± 0.0016mmol/mol lys and 0.0087 ± 0.0024 mmol/mol lys, respectively).The plasma protein content of pentosidine residues was increasedthree-fold in CRF patients, two-fold in PD patients, and four-foldin HD patients before dialysis (Figure 6g). In contrast, therewas no significant difference in the plasma protein contentof MOLD residues in uremic patients (Figure 6h). The plasmaprotein content of MetSO residues of control subjects was 0.78± 0.05 mmol/mol met and was increased approximately two-foldin CRF patients, PD patients, and HD patients before HD. Theplasma protein content of MetSO residues was not increased significantlyafter an HD session (Figure 4c). The plasma protein contentof 3-NT residues of control subjects was 0.0031 ± 0.0009mmol/mol tyr and was increased three-fold in PD patients andHD patients before HD. The plasma protein content of 3-NT residueswas decreased during HD (Figure 4d).
Figure 6. Concentrations of glycation adduct residues in plasma protein of normal control subjects, CRF patients, and ESRD patients with PD and HD therapy. (a) FL. (b) CML. (c) CEL. (d) G-H1. (e) MG-H1. (f) 3DG-H. (g) Pentosidine. (h) MOLD. CON, controls; PD (2), PD patients 2 h after change of PD fluid; HD (0) and HD (4), HD patients before and after a 4-h dialysis session, respectively. Data are mean ± SEM (n = 7 to 8). *P < 0.05, **P < 0.01, and ***P < 0.001, with respect to control subjects; oP < 0.05 with respect to CRF and PD patients; +P < 0.05 with respect to HD (0).
Plasma Concentrations of MCP-1 and TGF-
The plasma concentration of the chemokine MCP-1 was increasedsignificantly in PD and HD patients, with respect to controlsubjects (Figure 7A); estimates in HD patients after dialysiswere corrected for loss of plasma water by normalizing to albuminconcentration predialysis. The plasma concentration of solubleTGF- was increased in HD patients only, before and after a dialysissession, with respect to control subjects and CRF and PD patients(Figure 7B).
Figure 7. Plasma concentrations of monocyte chemotactic protein-1 (MCP-1) and TGF- of normal control subjects, CRF patients, and ESRD patients with PD and HD therapy. (a) MCP-1. (b) TGF-. CON, controls; PD (2), PD patients 2 h after change of PD fluid; HD (0) and HD (4), HD patients before and after a 4-h dialysis session, respectively. Data are mean ± SEM (n = 7 to 8). *P < 0.05 and **P < 0.01 with respect to control subjects. A correction factor using the individual changes in serum albumin concentrations before and after dialysis was applied to post-HD (4) data to allow for plasma water loss during HD (mean correction factor = 0.88).
Correlation Analysis of Protein Glycation, Oxidation, and Nitration Adducts
The relationships between glycation, oxidation, and nitrationadduct residues and free adducts in normal control subjectsand CRF, PD, and HD patients were examined. For HD patients,analyte values were predialysis estimates. Homotypic correlationsare correlations of glycation, oxidation, and nitration adductsof the same type in different forms or locations, namely plasmaprotein residues and free adducts in plasma and urine. Therewere no homotypic correlations in control subjects and CRF patients.In HD and PD patients, the plasma concentration of CML freeadduct correlated positively with the concentration of CML residues(r = 0.51, P < 0.05), and the plasma concentration of 3DG-Hfree adduct correlated positively with the concentration of3DG-H residues (r = 0.53, P < 0.05). Other correlation analysesof protein glycation, oxidation, and nitration adduct analytesand related variables are given for control subjects (Tables 5through 7), CRF patients (Tables 8 through 10), and PD andHD patients (Tables 11 and 12). These are provided, withoutdetailed explanation, for substantiation and interpretationin future studies.
In this study, we report the marked accumulation of proteinglycation free adducts in CRF and ESRD. Glycation free adductsnormally represent <5% of the total glycation adducts inplasma, but this increases substantially for some glycationadducts in uremia—particularly MG-H1. Overall, the totalconcentration of glycation free adducts in blood plasma increasedin the order control<CRF<PD<HD (predialysis) subjects.This pattern of high plasma concentrations of glycation freeadducts in uremic patients reflects marked decreased clearanceof glycation free adducts in CRF patients. Marked decreasedclearance was also generally found in PD patients, in whom theeffect was exacerbated by increased glycation free adduct formation—asindicated by increased glycation free adduct excretion. Theincreased plasma concentration of glycation free adducts inHD may also be due to increased formation and decreased clearanceof glycation free adducts.
In CRF patients, compared with creatinine, we found a disproportionateincrease in plasma concentration of glycation free adducts andprecipitous decline in their clearance while excretion rateswere normal—except for FL and pentosidine. This may indicatethat there is active tubular secretion of free adducts and thisis impaired in CRF. Some other uremic toxins were actively secretedin renal tubules by organic anion transporters (16). The plasmaconcentration of glycation free adducts correlated negativelywith creatinine clearance, but there was no correlation of glycationfree adduct clearances with molecular mass (data not shown).This is probably because glycation free adducts all have lowmolecular mass (<400 Da) and hence are freely filtered bythe glomerulus.
In PD therapy, excretion of protein glycation, oxidation, andnitration free adducts occurs mainly by elimination in the peritonealdialysate—except for pentosidine. D/P ratios were >1for most analytes (FL, CML, CEL, G-H1, MG-H1, argpyrimidine,and MetSO) at the end of the 4-h PET. Glycation and oxidationfree adducts are proposed to pass through the hypothesized small5-nm pores of micromolecular transport (17), as indeed do aminoacids (18), although additional active transport may enhancesolute entry into the peritoneal cavity; cf the role of aquaporin-1in the transport of water (19). The probable explanations forglycation and oxidation free adduct concentrations in the peritonealdialysate being higher than in plasma are (1) formation of glycationand oxidation free adducts in the peritoneal cavity by glycationand oxidation of amino acids and matrix proteins (with subsequentproteolysis) by glucose degradation products in the PD fluidand (2) active transport of glycation and oxidation free adductsacross the endothelium of peritoneal capillaries and vessels.Increased FL (as indicated by the assay of furosine), 3DG-H,other AGE, and oxidative markers in protein of peritoneal dialysate,vascular wall, and mesothelium have been reported (20–24)and were decreased by use of low glucose degradation productsthat contained PD fluid (24). Glycation and oxidation free adductsprobably utilize amino acid transporters to cross the capillaryendothelium of the peritoneum (25). This transport may contributeto amino acid loss into the peritoneal dialysate in PD therapy(18). The MTAC value for creatinine was similar but pentosidinewas higher than reported values (20); the MTAC for 3DG-H waslower than that of creatinine.
The excretion of most protein glycation, oxidation, and nitrationfree adducts was increased in PD patients, with excretions ofMetSO and MG-H1 exceptionally high. This may reflect increasedprotein modification in carbonyl, oxidative, and nitrosativestress of uremia. With reduced functional kidney mass and eliminationvia the peritoneal cavity, MetSO escapes metabolism by renalMetSO reductase (26); hence, the excretion rate and clearanceof MetSO were increased markedly (16-fold). The nine-fold increasein excretion of MG-H1 in PD patients was the most marked increasefound for the excretion of a glycation free adduct. Increasedprotein glycation by MG in ESRD is expected because of the highconcentrations of MG found in the plasma of ESRD patients andalso in PD fluids (9–11,27).
HD therapy of ESRD patients removed glycation free adducts fromthe circulation by passage through the dialysis membrane. Proteinglycation and oxidation free adducts have molecular masses <400Da and therefore are small molecular mass uremic toxins (2).At the end of the dialysis session, protein glycation free adductswere decreased equally in both plasma and hemodialysate butnot normalized to control levels—except for pentosidineand MOLD. RR values of glycation free adducts ranged from 39to 84%, demonstrating heterogeneity of clearance, and were lowerthan urea for only FL, CEL, and G-H1. Increased frequency ofHD with higher flux may improve the elimination of glycationfree adducts.
Most protein glycation, oxidation, and nitration adduct moietiesin plasma are modified amino acid residues of plasma proteinand are not eliminated efficiently by PD and HD therapy. Theglycation adduct residue content of plasma protein may be affectedby hypoalbuminemia in renal failure and related changes in albuminturnover (Table 1). Hypoalbuminemia in PD patients is causedby loss of albumin in urine and peritoneal dialysate and iscountered by increased albumin synthesis—inflammationand poor nutrition limit this response (28). This leads to adecreased half-life of albumin. Glycation of proteins in CRFand ESRD is expected to increase because of increased concentrationsof -oxoaldehyde glycating agents (9,10,27). The concentrationof FL residues may be decreased in CRF and HD patients by increasedoxidative stress with conversion of FL to CML residues (15).High levels of glycoxidation adducts (pentosidine and CML) inHD may reflect increased oxidative stress, relative to PD (29).The concentrations of MG-H1, MetSO, and 3-NT residues were normalizedin HD patients after dialysis. 3-NT–modified albumin undergoespreferential endothelial transcytosis (30), which may be increasedin the inflammatory response to dialysis. MetSO residues maybe reduced to methionine residues by MetSO reductase duringendothelial transcytosis. MG-H1–modified proteins maybe bound by activated monocytes in the dialysis session anddegraded (31). The concentration of MG-H1 residues in plasmaprotein of PD and HD patients correlated negatively with theplasma concentration of albumin (Table 11). Decreased albuminconcentration is associated with poor survival of ESRD patients(23).
The few homotypic correlations of glycation adduct protein residuesand free adducts reflect important nonvascular sources of glycationfree adduct formation: Degradation of glycated tissue proteins(4), adsorption from digested food protein (8), and glycationof amino acids (13). Major sources of glycation and oxidationfree adducts are likely to be cellular proteolysis and digestedfood (4,32).
The plasma concentration of MCP-1 was increased significantlyin PD and HD patients, and the plasma concentration of solubleTGF- was increased in HD patients only, confirming previousreports (33–36). MCP-1 is associated with macrophage infiltrationinto the peritoneal cavity of PD patients and correlates withsoluble adhesion molecule expression in HD (33,37). TGF- isa key mediator of renal fibrosis in progressive renal failureand ESRD. The concentration of MG-H1 and 3DG-H residues in plasmaprotein correlated positively with TGF- in CRF patients. Increasedexpression of TGF- in CRF is associated with a progressive declineof residual renal function and increased fibrosis leading toESRD. Although protein glycation, oxidation, and nitration freeadducts are eliminated by dialysis, plasma MCP-1 and TGF- concentrationswere maintained by dialysis therapy. The inflammation of dialysisand the increase AGE may be linked, but there is no direct relationshipwith the clearance of AGE free adducts.
Table 9 Correlation of protein glycation, oxidation, and nitration free adducts in plasma in CRF patients
Acknowledgments
We thank Baxter Healthcare Inc (Deerfield, IL) for support fora PhD studentship for S.A. and the Wellcome Trust for supportfor our protein biomarker research. We thank the staff and patientsof the Renal Unit, St. Bartholomew's Hospital, for their participationin this study.
Footnotes
Published online ahead of print. Publication date availableat www.jasn.org.
Thornalley PJ, Battah S, Ahmed N, Karachalias N, Agalou S, Babaei-Jadidi R, Dawnay A: Quantitative screening of advanced glycation endproducts in cellular and extracellular proteins by tandem mass spectrometry.
Biochem J 375
: 581
–592, 2003[CrossRef][Medline]
Wells-Knecht MC, Lyons TJ, McCance DR, Thorpe SR, Baynes JW: Age-dependent increases in ortho-tyrosine and methionine sulfoxide in human skin collagen is not accelerated in diabetes.
J Clin Invest 100
: 839
–846, 1997[Medline]
Gaut JP, Byun J, Tran HD, Heinecke JW: Artifact-free quantitation of free 3-chlorotyrosine, 3-bromotyrosine, and 3-nitrotyrosine in human plasma by electron capture-negative chemical ionization gas chromatography mass spectrometry and liquid chromatography-electrospray ionization tandem mass spectrometry.
Anal Biochem 300
: 252
–259, 2002[CrossRef][Medline]
Faist V, Erbersdobler H: Metabolic transit and in vivo effects of melanoidins and precursor compounds deriving from the Maillard reaction.
Ann Nutr Met 45
: 1
–12, 2001
Henle T: AGEs in foods: Do they play a role in uremia?
Kidney Int 63
: S145
–S147, 2003[CrossRef]
Odani H, Shinzato T, Matsumoto Y, Usami J, Maeda K: Increase in three alpha,beta-dicarbonyl compound levels in human uremic plasma: Specific in vivo determination of intermediates in advanced Maillard reaction.
Biochem Biophys Res Commun 256
: 89
–93, 1999[CrossRef][Medline]
Agalou S, Karachalias N, Tucker B, Thornalley PJ, Dawnay A: Estimation of alpha-oxoaldehydes formed from the degradation of glycolytic intermediates and glucose fragmentation in blood plasma of human subjects with uraemia.
Int Congr Ser 1245
: 182
–93, 2002
Wieslander A, Linden T, Musi B, Jarkelid L, Speidel R, Beck W, Henle T, Deppisch R: Exogenous uptake of carbonyl stress compounds promoting AGE formation from peritoneal dialysis fluids.
Contrib Nephrol 131
: 82
–89, 2001
Degenhardt TP, Thorpe SR, Baynes JW: Chemical modification of proteins by methylglyoxal.
Cell Mol Biol 44
: 1139
–1145, 1998[Medline]
Ahmed N, Argirov OK, Minhas HS, Cordeiro CA, Thornalley PJ: Assay of advanced glycation endproducts (AGEs): Surveying AGEs by chromatographic assay with derivatisation by aminoquinolyl-N-hydroxysuccimidyl-carbamate and application to Nepsilon-carboxymethyl-lysine- and Nepsilon-(1-carboxyethyl)lysine-modified albumin.
Biochem J 364
: 1
–14, 2002[Medline]
Krediet RT, Lindholm B, Rippe B: Pathophysiology of peritoneal membrane failure.
Perit Dial Int 20
: S22
–S42, 2000[Free Full Text]
Smith PR, Thornalley PJ: Mechanism of the degradation of non-enzymatically glycated proteins under physiological conditions. (Studies with the model fructosamine, N-(1-deoxy-D-fructose-1-yl)hippuryl-lysine).
Eur J Biochem 210
: 729
–739, 1992[Medline]
Deguchi T, Kusuhara H, Takadate A, Endou H, Otagiri M, Sugiyama Y: Characterization of uremic toxin transport by organic anion transporters in the kidney.
Kidney Int 65
: 162
–174, 2004[CrossRef][Medline]
Ronco C, Brendolan A, La Greca G: The peritoneal dialysis system.
Nephrol Dial Transplant 13
: 94
–99, 1998[Free Full Text]
de la Motte S, Plum J, Passlick-Deetjen J, Grabensee B: One-compartment model for amino acids and other biological molecules in peritoneal dialysis.
Int J Clin Pharmacol Ther 40
: 60
–68, 2002[Medline]
Ruiz MC, Portero-Otin M, Pamplona R, Requena JR, Prat J, Lafarga MA, Borras M, Bellmunt MJ: Chemical and immunological characterization of oxidative nonenzymatic protein modifications in dialysis fluids.
Perit Dial Int 23
: 23
–32, 2003[Abstract/Free Full Text]
Friedlander MA, Wu YW, Elgawish A, Monnier VM: Early and advanced glycosylation end products. Kinetics of formation and clearance in peritoneal dialysis.
J Clin Invest 97
: 728
–735, 1996[Medline]
Nakamura S, Miyazaki S, Sakai S, Morita T, Hirasawa Y, Niwa T: Localization of imidazolone in the peritoneum of CAPD patients: A factor for a loss of ultrafiltration.
Am J Kidney Dis 38
: S107
–S110, 2001[Medline]
Honda K, Nitta K, Horita S, Yumura W, Nihei H, Nagai R, Ikeda K, Horiuchi S: Accumulation of advanced glycation end products in the peritoneal vasculature of continuous ambulatory peritoneal dialysis patients with low ultra-filtration.
Nephrol Dial Transplant 14
: 1541
–1549, 1999[Abstract/Free Full Text]
Owen WF, Lew NL, Liu Y, Lowrie EG, Lazarus JM: The urea reduction ratio and serum-albumin concentration as predictors of mortality in patients undergoing hemodialysis.
N Engl J Med 329
: 1001
–1006, 1993[Abstract/Free Full Text]
Mortier S, Faict D, Schalkwijk CG, Lameire NH, De Vriese AS: Long-term exposure to new peritoneal dialysis solutions: Effects on the peritoneal membrane.
Kidney Int 66
: 1257
–1265, 2004[CrossRef][Medline]
Mann GE, Yudilevich DL, Sobrevia L: Regulation of amino acid and glucose transporters in endothelial and smooth muscle cells.
Physiol Rev 83
: 183
–252, 2003[Abstract/Free Full Text]
Weissbach H, Etienne F, Hoshi THSH, Lowther WT, Matthews B, St. John G, Nathan C, Brot N: Peptide methionine sulfoxide reductase: Structure, mechanism of action, and biological function.
Arch Biochem Biophys 397
: 172
–178, 2002[CrossRef][Medline]
Mann VM, Tucker B, Thornalley PJ, Dawnay A: Elevated plasma methylglyoxal and glyoxal in uraemia: Implications for advanced glycation endproduct formation.
Kidney Int 55
: 2582
–178, 1999
Kaysen GA: Albumin turnover in renal disease.
Miner Electrolyte Metab 24
: 55
–63, 1998[CrossRef][Medline]
Lim PS, Chang YM, Thien LM, Wang NP, Yang CC, Chen T, Hsu WM: 8-Iso-prostaglandin F-2 alpha as a useful clinical biomarker of oxidative stress in ESRD patients.
Blood Purif 20
: 537
–542, 2002[CrossRef][Medline]
Predescu D, Predescu S, Malik AB: Transport of nitrated albumin across continuous vascular endothelium.
Proc Natl Acad Sci USA 99
: 13932
–13937, 2002[Abstract/Free Full Text]
Westwood ME, Argirov OK, Abordo EA, Thornalley PJ: Methylglyoxal-modified arginine residues—A signal for receptor-mediated endocytosis and degradation of proteins by monocytic THP-1 cells.
Biochim Biophys Acta 1356
: 84
–94, 1997[Medline]
Liardon R, de Weck-Gaudard D, Philipossian G, Finot P-A: Identification of Nespsilon-carboxymethyllysine: A new Maillard reaction product, in rat urine.
J Agric Food Chem 35
: 427
–431, 1987[CrossRef]
Jacobson SH, Thylen P, Lundahl J: Three monocyte-related determinants of atherosclerosis in haemodialysis.
Nephrol Dial Transplant 15
: 1414
–1419, 2000[Abstract/Free Full Text]
Jacobson SH, Hylander B, Thylen P, Lundahl J: Monocyte-related determinants of inflammation in patients on peritoneal dialysis.
Am J Nephrol 21
: 40
–46, 2001[CrossRef][Medline]
Fujisawa M, Haramaki R, Miyazaki H, Imaizumi T, Okuda S: Role of lipoprotein (a) and TGF-beta 1 in atherosclerosis of hemodialysis patients.
J Am Soc Nephrol 11
: 1889
–1895, 2000[Abstract/Free Full Text]
Stompor T, Zdzienicka A, Motyka M, Dembinska-Kiec A, Davies SJ, Sulowicz W: Selected growth factors in peritoneal dialysis: Their relationship to markers of inflammation, dialysis adequacy, residual renal function, and peritoneal membrane transport.
Perit Dial Int 22
: 670
–676, 2002[Abstract/Free Full Text]
Tekstra J, Visser CE, Tuk CW, BrouwerSteenbergen JE, Burger CW, Krediet RT, Beelen RHJ: Identification of the major chemokines that regulate cell influxes in peritoneal dialysis patients.
J Am Soc Nephrol 7
: 2379
–2384, 1996[Abstract]
Received for publication August 4, 2004.
Accepted for publication February 16, 2005.
This article has been cited by other articles:
H. Vlassara, W. Cai, S. Goodman, R. Pyzik, A. Yong, X. Chen, L. Zhu, T. Neade, M. Beeri, J. M. Silverman, et al. Protection against Loss of Innate Defenses in Adulthood by Low Advanced Glycation End Products (AGE) Intake: Role of the Antiinflammatory AGE Receptor-1
J. Clin. Endocrinol. Metab.,
November 1, 2009;
94(11):
4483 - 4491.
[Abstract][Full Text][PDF]
E. G. Gerrits, A. J. Smit, and H. J. G. Bilo AGEs, autofluorescence and renal function
Nephrol. Dial. Transplant.,
March 1, 2009;
24(3):
710 - 713.
[Full Text][PDF]
N. Rabbani and P. J. Thornalley QUANTITATION OF MARKERS OF PROTEIN DAMAGE BY GLYCATION, OXIDATION, AND NITRATION IN PERITONEAL DIALYSIS
Perit. Dial. Int.,
February 1, 2009;
29(Supplement_2):
S51 - S56.
[Abstract][Full Text][PDF]
Y. Wang, W. Beck, R. Deppisch, S. M. Marshall, N. A. Hoenich, and M. G. Thompson Differential effects of dialysis and ultrafiltrate from individuals with CKD, with or without diabetes, on platelet phosphatidylserine externalization
Am J Physiol Renal Physiol,
January 1, 2008;
294(1):
F220 - F228.
[Abstract][Full Text][PDF]
G. Cohen, G. Glorieux, P. Thornalley, E. Schepers, N. Meert, J. Jankowski, V. Jankowski, A. Argiles, B. Anderstam, P. Brunet, et al. Review on uraemic toxins III: recommendations for handling uraemic retention solutes in vitro towards a standardized approach for research on uraemia
Nephrol. Dial. Transplant.,
December 1, 2007;
22(12):
3381 - 3390.
[Full Text][PDF]
S. Bengmark Advanced Glycation and Lipoxidation End Products-Amplifiers of Inflammation: The Role of Food
JPEN J Parenter Enteral Nutr,
September 1, 2007;
31(5):
430 - 440.
[Abstract][Full Text][PDF]
C. P. Schmitt, D. von Heyl, S. Rieger, K. Arbeiter, K. E. Bonzel, M. Fischbach, J. Misselwitz, A.-K. Pieper, F. Schaefer, and for the Mid European Pediatric Peritoneal Dialysis Reduced systemic advanced glycation end products in children receiving peritoneal dialysis with low glucose degradation product content
Nephrol. Dial. Transplant.,
July 1, 2007;
22(7):
2038 - 2044.
[Abstract][Full Text][PDF]
F. Galli Protein damage and inflammation in uraemia and dialysis patients
Nephrol. Dial. Transplant.,
July 1, 2007;
22(suppl_5):
v20 - v36.
[Abstract][Full Text][PDF]
L. Segall and A. Covic Cardiovascular disease in haemodialysis and peritoneal dialysis: arguments pro haemodialysis
Nephrol. Dial. Transplant.,
January 1, 2007;
22(1):
59 - 63.
[Full Text][PDF]
J. M. Bargman Peritoneal dialysis solutions and patient survival: does wishing make it so?
Nephrol. Dial. Transplant.,
October 1, 2006;
21(10):
2684 - 2686.
[Full Text][PDF]
Top
Abstract
Introduction
-carboxymethyl-lysine (CML), N
-oxoaldehydes, particularly glyoxal, MG, and 3-DG (9,10). 
10%, and analyte estimates corroborated with independent measurements when these were available (6,12). Free glycation adducts were determined by assay of analytes in ultrafiltrate (12 kD filter cutoff, 50-µl aliquot) of plasma, urine, and dialysate. Glycation adduct residues of plasma protein were determined in exhaustive enzymatic digests (50-µg protein equivalent) (4,13). Routine clinical service methods were used for urine and plasma creatinine (Jaffe rate method) and serum albumin (BCG colorimetric assay) on an Olympus analyzer. Plasma monocyte chemotactic protein-1 (MCP-1) and TGF-
were assayed in platelet free plasma by double-antibody ELISA (R & D Systems, Abingdon, UK). 
statistic. Mass transfer area coefficients (MTAC) were deduced by the simplified Garred equation (14). 
