Advanced Glycation End Products and Inflammatory Cytokine Profiles in Maintenance Hemodialysis Patients After the Ingestion of a Protein-Dense Meal

Objective: The goal of this investigation was to evaluate circulating and skeletal muscle inflammatory biomarkers between maintenance hemodialysis (MHD) and demographic-matched control subjects (CON) before and after ingestion of a protein-rich meal. Design and Methods: CON (n = 8; 50 ± 2 years; 31 ± 1 kg/m2) and MHD patients (n = 8; 56 ± 5 years; 32 ± 2 kg/m2) underwent a basal blood draw and muscle biopsy and serial blood draws after the ingestion of a mixed meal on a nondialysis day. Plasma advanced glycation end products (AGEs) and markers of oxidation were assessed via liquid chromatography–tandem mass spectrometry before and after the meal (+240 min). Circulating inflammatory cytokines and soluble receptors for AGE (sRAGE) isoforms (endogenous secretory RAGEs and cleaved RAGEs) were determined before and after the meal (+240 min). Basal muscle was probed for inflammatory cytokines and protein expression of related signaling components (RAGE, Toll-like receptor 4, oligosaccharyltransferase subunit 48, TIR-domain–containing adapter-inducing interferon-β, total IκBα, and pIκBα). Results: Basal circulating AGEs were 7- to 343-fold higher (P < .001) in MHD than those in CON, but only MG-H1 increased in CON after the meal (P < .001). There was a group effect (MHD > CON) for total sRAGEs (P = .02) and endogenous secretory RAGEs (P < .001) and a trend for cleaved RAGEs (P=.09), with no meal effect. In addition, there was a group effect (MHD < CON; P < .05) for circulating fractalkine, interleukin (IL)10, IL17A, and IL1β and a trend (P < .10) for IL6 and macrophage inflammatory protein 1 alpha, whereas tumor necrosis factor alpha was higher in MHD (P < .001). In muscle, Toll-like receptor 4 (P = .03), TIR-domain–containing adapter-inducing interferon-β (P = .002), and oligosaccharyltransferase subunit 48 (P = .02) expression was lower in MHD than that in CON, whereas IL6 was higher (P = .01) and IL8 (P = .08) tended to be higher in MHD. Conclusion: Overall, MHD exhibited an exaggerated, circulating, and skeletal muscle inflammatory biomarker environment, and the meal did not appreciably affect the inflammatory status.


Introduction
2][3] Chronically elevated circulating advanced glycation end products (AGEs) are a major driver of inflammation in patients with chronic kidney disease (CKD), [4][5][6][7][8] and progression along the CKD continuum is related to increased circulating AGE load. 8,94][15][16] This ligand-receptor system can be disrupted through the proteolytic release of RAGE, forming a cleaved RAGE isoform (cRAGE) 17 and alternative RAGE splicing, leading to the production and cellular expulsion of endogenous secretory RAGE (esRAGE). 18cRAGE and esRAGE (total sRAGE) suppress AGE-RAGE signaling by (1) reducing signaling potential at the cell due to fewer RAGE receptors and (2) acting as a decoy through binding and sequestering circulating AGEs. 70][21] In support of this link between inflammation and proteostasis in MHD, past efforts have shown a relationship between release of the proinflammatory cytokine interleukin-6 (IL6) from peripheral tissues and accelerated muscle protein loss. 21urthermore, markers of circulating inflammation, such as c-reactive protein (CRP), are strongly related to indirect measurements of skeletal muscle protein synthesis, degradation, and net protein balance in MHD patients. 19Proinflammatory cytokines exert their effects on protein balance by activating downstream proteolytic signaling pathways, including the ubiquitin-proteasome pathway and caspase-3.We recently demonstrated that individuals undergoing MHD have higher basal caspase-3 protein content in skeletal muscle. 22ctivation of pathways that regulate protein breakdown rates leads to the excess release of amino acids into the muscle intracellular free pool, resulting in elevated muscle protein synthesis rates in the basal state. 23This overstimulation of MHD muscle results in anabolic resistance of muscle protein synthesis rates to dietary protein and, over time, likely contributes to accelerated losses in muscle mass in individuals on MHD. 23reover, protein-energy wasting (i.e., depletion of body protein stores) is another hallmark of advanced CKD and likely contributes to the progression of muscle loss with MHD.5][26][27] Evidence exists showing MHD treatment is catabolic in nature, 28 an outcome potentially driven by modulation of genes involved with apoptosis and inflammation in skeletal muscle 29 and loss of amino acids in the dialysate. 30Coupled with the upregulation of inflammatory signaling pathways in muscle, loss of amino acids during MHD highlights the importance of consuming protein-rich meals between dialysis treatments to counteract the catabolic nature of MHD.
Despite continued illumination of the underlying mechanisms promoting inflammation, our understanding of the inflammatory phenotype in MHD is still limited.It is clear that patients on MHD experience muscle wasting, a condition suspected to be perpetuated by chronic basal inflammation and poor nutrient status.A more complete understanding of the inflammatory burden in patients undergoing MHD will direct targeted therapeutics.Thus, the goal of this investigation was to evaluate the spatial and temporal dynamics of inflammation and inflammatory regulators by evaluating a host of biomarkers in the circulation and skeletal muscle in the basal state and following a mixed, protein-dense meal recommended by the National Kidney Foundation for an MHD population 31 and a demographic-matched control group.Furthermore, we aimed to build on our previous work and explore a potential relationship between the inflammatory milieu and anabolic resistance to feeding observed in this MHD cohort. 22

Experimental Design
Recruitment methods for this pilot study have been previously reported, representing a sub-analysis of this work. 22Briefly, volunteers were determined to be eligible based on responses to a detailed medical screening questionnaire and blood panel findings.In addition, MHD participants received clearance to participate in the study from their personal nephrologist.All participants were informed about the experimental procedures and potential risks before providing written consent.The study was approved by the Institutional Review Board at the University of Illinois at Urbana-Champagne and conformed to standards for the use of human participants in research as outlined in the Declaration of Helsinki.Clinical Trial Registration: (www.clincialtrials.gov/).

Protocol
A detailed description of the protocol has been described elsewhere. 22Briefly, all participants were instructed to refrain from vigorous physical activity for 3 days before the trial and maintain their normal dietary pattern and prescribed medications.Subjects reported to the laboratory at 7:00 AM after consuming a 320 kcal standardized meal (22 g protein, 43 g carbohydrate, and 7 g fat) the night before.Other than this standardized meal, participants were instructed to remain fasted.MHD patients were studied ~24 h after their previous dialysis treatment.The study protocol consisted of a basal blood draw and skeletal muscle biopsy followed by serial blood draws (30, 60, 90, 120, 180, 240, and 300 minutes) after a mixed meal challenge.Circulating postmeal inflammatory assessments were made at the 240-minute time point.Coupled with sample availability, our intention was to capture the inflammatory load in both the postabsorptive and postprandial states.In addition, all participants performed a 2-day dietary recall (Nutritionist Pro v2.1.13,Axxya Systems, Redmond, WA).On average, protein intake tended to be lower (p < .10) in MHD than that in CON (MHD: 0.9 ± 0.14; CON: 1.14 ± 0.05 g/kg body weight).No differences in other macronutrients were observed (p > .05).

Mixed Meal Challenge
After basal measurements, each participant consumed a mixed meal consisting of 3 scrambled eggs, 1 slice of toasted white bread, 300 mL apple juice, and 10 g cow butter (total: 546 kcal, 20 g protein, 59 g carbohydrate, and 26 g fat).This meal was selected because it is representative of a standard meal recommended by the National Kidney Foundation for this patient population. 31Macronutrient composition of the eggs was determined by the combustion method (method 990.03;AOAC International, 2000; TruMac; LEO Corp., Sain Joseph, MI).For the other foods, macronutrient composition was determined from their respective food labels.

Blood Draws and Skeletal Muscle Biopsy
A catheter was inserted into a dorsal hand vein and heated for sampling of arterialized venous blood.A skeletal muscle biopsy [32][33][34][35] of the m.vastus lateralis was obtained in the basal state under local anesthetic (Lidocaine, 2%).After the muscle biopsy, excess blood, visible fat, and connective tissue were removed and the sample was immediately flash-frozen in liquid nitrogen and stored at −80°C until analysis.
Muscle inflammatory targets (fractalkine, interferon gamma, IL10, IL17 A, IL1β, IL6, IL8, MIP1α, and TNFα) were determined by Milliplex magnetic bead assay as described previously.Skeletal muscle samples were homogenized by ceramic beads (Lysing Matrix D; MP Biomedical; Irvine, CA) in 20 volumes of ice-cold PBS supplemented with 1X protease and phosphatase inhibitors (5872S, Cell Signaling Technology).After bead homogenization, samples were sonicated 2 × 30 seconds at 50% power on ice.Sonicated samples were then spun at 20000g for 30 minutes at 4°C, and the supernatant was retained for total protein concentration assessment via bicinchoninic acid assay (Pierce Biotechnology, Rockford, IL) and cytokine analyses.Antibody beads, controls, buffers, matrix (prepared in the muscle homogenization buffer), and standards were prepared, and the kit was run as per manufacturers' guidelines.

Statistical Analyses
Data were analyzed with GraphPad Prism (GraphPad Software, La Jolla, CA, USA) and evaluated for normality with a Shapiro-Wilk test.Non-normally distributed data were log-transformed before statistical analyses.A two-way analysis of variance with repeated measures was used to compare circulating OX markers, AGEs, inflammatory factors, and sRAGE concentrations between MHD and CON and time points (basal and 240 minutes after meal).Tukey's post hoc test was used to examine specific differences when appropriate.Student's t-test was used to compare subject characteristics, changes (absolute and percent) in circulating inflammatory concentrations, and basal inflammatory muscle protein expression between groups.Relationships between variables of interest were analyzed via Spearman's rho correlation.Significance was set at P < .05,and a trend toward significance was recognized as P < .10.Data are presented as mean ± SE.

Circulating Markers of Protein OX and AGEs
Plasma protein OX and AGEs in the basal and postprandial state are presented in Figure 1.Analysis of plasma OX yielded no basal differences (P > .05) in methionine sulfoxide or AAA between MHD and CON.An effect of time was observed for methionine sulfoxide (P = .03)and AAA (P <.001).For the circulating AGEs, basal concentrations were higher (P < .001) in MHD than those in CON (range = 7-to 343-fold greater).After the meal, CEL, CML, 3DG-H, and G-H1 remained unchanged (P > .05) in both groups, whereas MG-H1 increased 11.1-fold (P < .001) at 240 minutes in CON.Although MG-H1 increased in CON, concentrations remained ~28 fold lower than MHD at 240 minutes of the postprandial period.

Muscle Protein Expression
Basal receptor signaling component protein expression in skeletal muscle of MHD and CON participants determined via Western blot is presented in Figure 3. Contrary to our original hypotheses, TLR4 (−27%; P = .03),TRIF (−74%; P = .002),and OST48 (−55%; P = .02)expression was lower in MHD than that in CON, whereas the RAGE and pIκBα/total IκBα were similar between groups (P > .05).To complement the inflammatory receptor protein expression, a targeted panel of nine inflammatory cytokines was also evaluated (Figure 4).Findings (pg target protein) were normalized in three ways: pg/mg muscle wet wt (Figure 4), pg/mL sample homogenate analyzed, and pg/μg total protein.Subtle statistical differences were observed among the three normalization approaches (see Table S1 for pg/mL and pg/μg protein normalizations).When cytokine protein was normalized to mg muscle wet weight, basal IL6 (+141%; P = .01)and IL8 (+70%; P = .08)were higher in MHD than those in CON, whereas the other cytokines were similar (P > .05).

Discussion
The overarching goal of this investigation was to provide insight into the inflammatory phenotype in MHD by evaluating a comprehensive, yet targeted, group of circulating and skeletal muscle inflammatory biomarkers in the basal state and after a typical meal rich in protein recommended for this population.The principal findings from this study include the following: (1) circulating AGEs are substantially elevated in MHD patients compared with CON, (2) the basal circulating and skeletal muscle cytokine profile in MHD is generally proinflammatory, and (3) there was minimal effect of meal ingestion on modulating all inflammatory biomarkers in both groups.
A major objective of hemodialysis is to remove uremic solutes that exert toxic effects.Using the gold-standard quantification methods, 42 we report that basal plasma AGE concentrations are 7-to 343-fold higher in the MHD patients than those in demographic controls.
Noteworthy is that these assessments were made on samples from a nondialysis day (~24 hours after finishing dialysis), which suggests chronic AGE toxicity in MHD patients rather than only acutely during the dialysis period.Our results are consistent with previous reports, 4,43,44 and note that MG-H1 is the most abundant AGE in MHD patients.4][45] Interestingly, the lack of correlation between reduction in AGE levels and AGE molecular size with dialysis 43 highlights factors other than filter pore size govern AGE removal; thus, modified filtration approaches may need to be reconsidered.
4][15][16] Interestingly, RAGE expression was similar between groups, and OST48, TLR4, and TRIF expression was lower in MHD than that in the control group.This is intriguing and, to the best of our knowledge, the first report of RAGE protein expression in skeletal muscle of individuals undergoing MHD.In contrast to the findings presented here, one of the few studies conducted on circulating human cells demonstrated elevated RAGE protein expression in peripheral blood monocytes of dialysis patients, 46 and another reported higher peripheral blood monocyte RAGE mRNA expression in patients with advanced CKD. 47 Findings of high RAGE expression in circulating cells and similar RAGE expression in patients with renal disease suggest a potential tissue-specific effect.Interestingly, OST48 expression was lower in MHD.OST48 is a membrane-bound receptor that binds and transports AGEs from the extracellular to intracellular compartments for processing and removal. 48,49Therefore, OST48 may compete with RAGEs for ligands and likely exerts overall anti-inflammatory effects.In support, we observed a strong relationship between muscle OST48 and muscle IL10 (r = 0.59; P = .02).MHD muscle containing less OST48 than control subjects may be a fundamental component of the generally proinflammatory environment.
TLR4 is tasked with recognizing a host of ligands (some of which are shared with RAGEs) and signal to downstream pathway elements, including the TRIF adapter and the NFκB pathway.Similar to our work presented here in muscle, Kuroki et al. found reduced TLR4 expression in circulating cells compared with controls, and dialysis duration (in years) was inversely related to TLR4 expression. 50Other studies have shown no difference 51 or greater TLR4 expression in circulating cells of dialysis patients. 52To our knowledge, the only other study to examine TLR4 in the muscle of this patient population found higher TLR4 expression in dialysis patients than in a control group. 53The differential findings between our study and those by Verzola et al. may be explained by the analytical approach used to measure TLR4 (Western blot in our study vs. immunohistochemistry by Verzola et al.) or the type of dialysis technique used by the patients (i.e., MHD in our study and peritoneal dialysis by Verzola et al.).This is important to consider as hemodialysis is inflammatory in nature and has been shown to modulate genes involved with apoptosis and inflammation in skeletal muscle. 295][56] Chronic sRAGE treatment in db/db mice reduces glomerulosclerosis and improves renal function. 54In humans, sRAGE is strongly associated with muscle mass, 56 low sRAGE is related to the development of obesity and insulin resistance across the glucose tolerance continuum, 37 and weight loss is related to increased sRAGE. 57In line with other studies on patients undergoing dialysis, 55,58 here, we report elevated total sRAGE and esRAGE in MHD patients.One explanation for higher sRAGE in our MHD patients is the reliance on the kidney for sRAGE clearance, as total sRAGE concentration decreases after kidney transplantation. 59Another potential explanation for exaggerated sRAGE levels in clinical patients is this serves as a protective countermeasure to chronic inflammation.Therefore, in patients characterized by a proinflammatory environment, high sRAGE may be indicative of underlying pathologies.In fact, a very high sRAGE is associated with cardiovascular disease and all-cause mortality in clinical populations. 60It is clear that the sRAGE has protective effects; however, it remains to be seen if the sRAGE is upregulated as a compensatory mechanism to combat the inflammatory load in MHD or if impaired renal function blunts sRAGE clearance.
2,53 Overall, here, we show the basal profile in MHD is proinflammatory.In the circulation, MHD had more proinflammatory CRP and TNFα and less anti-inflammatory IL10 than the control group.CRP is a well-studied inflammatory biomarker and is linked to negative forearm protein balance, 19 often used as a proxy for skeletal muscle, and muscle loss 61 in dialysis patients.CRP has also been shown to be localized to atherosclerotic lesions in humans, and CRP treatment of human monocytes leads to TNFα production. 62Interestingly, we found a strong correlation between circulating CRP and TNFα concentration (r = 0.70; P = .004).TNFα drives muscle wasting through its activation of the NFκB pathway 63 and caspase 3. 64 Greater TNFα concentration in MHD muscle may explain the higher caspase 3 expression and hyperstimulation of myofibrillar protein synthesis rates in MHD reported in our previous publication. 22ovided muscle is a predominant metabolic tissue that produces and secretes inflammatory factors 65 and experiences substantial wasting in renal failure, 66 it is surprising that so little is known about inflammation concentrations in MHD muscle.In this study, we report a modest proinflammatory profile in MHD muscle as concentrations of the proinflammatory IL6 and chemotactic protein IL8 are elevated compared with the CON group.IL6 modulates its inflammatory effects in the muscle through protein balance disruption.In support, shortterm IL6 infusion into healthy humans reduces muscle protein turnover by 50%, 67 and IL6 injection into animals is reported to induce muscle wasting, and this effect is mediated by the suppressor of cytokine signaling 3. 68 Suppressor of cytokine signaling 3 is generally considered proteolytic as it upregulates caspase 3. Therefore, elevated IL6 in MHD muscle may provide additional insight into greater caspase 3 protein content previously reported in this MHD cohort. 22 summary, this study provides a comprehensive, yet targeted, analysis of circulating and skeletal muscle inflammatory biomarkers in MHD patients in the basal state and after the ingestion of a meal rich in protein.Our findings of higher AGEs and proinflammatory cytokines and lower OST48 and anti-inflammatory cytokines add to the growing literature, indicating an overall proinflammatory environment in patients undergoing chronic MHD.Given the impact of renal replacement therapy on appetite and nutrient availability, 20,[24][25][26][27] we also sought to explore the biomarker effect of a mixed meal.In general, there was minimal effect of the meal on biomarkers explored, indicating a typical meal for this population does not exacerbate or attenuate the pre-existing inflammatory load.These findings suggest that the overall patient care may need to evolve to generate more specific and complementary therapeutic approaches in patients with CKD.The overarching goal should emphasize strategies that reduce the inflammatory load, including tailored physical activity and nutritional programs, thereby alleviating anabolic resistance.

Practical Applications
In this study, we report an exaggerated circulating and skeletal muscle inflammatory environment, and a protein-rich meal did not appreciably affect inflammatory status.To curtail the loss in muscle mass MHD patients typically experience, these findings suggest dialysis treatments may need to evolve.Plasma inflammatory biomarkers in the basal state and after meal (+240 min) assessed via multiplex.See Mixed Meal Challenge in the Methods for more details regarding mixed meal nutrition information.CON, control subjects (n = 8); IFNγ, interferon gamma; IL, interleukin; MHD, maintenance hemodialysis patients (n = 8); MIP1α, macrophage inflammatory protein 1 alpha; TNFα, tumor necrosis factor alpha; Data are mean ± SE; * Group effect: P < .05versus CON; † Group effect: P < .10versus CON.

Table 1 .
Clinical Characteristics of Control Subjects (CON) and Maintenance Hemodialysis Patients (MHD)BMI, body mass index; BUN, blood urea nitrogen; CRP, c-reactive protein; eGFR, estimated glomerular filtrate rate; HOMA-IR, homeostatic model of assessment for insulin resistance.