- Split View
-
Views
-
Cite
Cite
Laetitia Koppe, Pascaline M Alix, Marine L Croze, Stéphane Chambert, Raymond Vanholder, Griet Glorieux, Denis Fouque, Christophe O Soulage, p-Cresyl glucuronide is a major metabolite of p-cresol in mouse: in contrast to p-cresyl sulphate, p-cresyl glucuronide fails to promote insulin resistance, Nephrology Dialysis Transplantation, Volume 32, Issue 12, December 2017, Pages 2000–2009, https://doi.org/10.1093/ndt/gfx089
- Share Icon Share
Abstract
The role of uraemic toxins in insulin resistance associated with chronic kidney disease (CKD) is gaining interest. p-Cresol has been defined as the intestinally generated precursor of the prototype protein-bound uraemic toxins p-cresyl sulphate (p-CS) as the main metabolite and, at a markedly lower concentration in humans, p-cresyl glucuronide (p-CG). The objective of the present study was to evaluate the metabolism of p-cresol in mice and to decipher the potential role of both conjugates of p-cresol on glucose metabolism.
p-CS and p-CG were measured by high performance liquid chromatography-fluorescence in serum from control, 5/6 nephrectomized mice and mice injected intraperitoneously with either p-cresol or p-CG. The insulin sensitivity in vivo was estimated by insulin tolerance test. The insulin pathway in the presence of p-cresol, p-CG and/or p-CS was further evaluated in vitro on C2C12 muscle cells by measuring insulin-stimulated glucose uptake and the insulin signalling pathway (protein kinase B, PKB/Akt) by western blot.
In contrast to in humans, where p-CS is the main metabolite of p-cresol, in CKD mice both conjugates accumulated, and after chronic p-cresol administration with equivalent concentrations but a substantial difference in protein binding (96% for p-CS and <6% for p-CG). p-CG exhibited no effect on insulin sensitivity in vivo or in vitro and no synergistic inhibiting effect in combination with p-CS.
The relative proportion of the two p-cresol conjugates, i.e. p-CS and p-CG, is similar in mouse, in contrast to humans, pinpointing major inter-species differences in endogenous metabolism. Biologically, the sulpho- (i.e. p-CS) but not the glucuro- (i.e. p-CG) conjugate promotes insulin resistance in CKD.
INTRODUCTION
Chronic kidney disease (CKD) is characterized by the gradual accumulation of retention solutes, some of which result from the degradation of dietary proteins. Recently, protein-bound uraemic solutes, derived from phenols and indoles, have been shown to interact with many biological systems. p-Cresol is mainly generated as an end-product of tyrosine biotransformation by anaerobic intestinal bacteria. It appeared that not p-cresol, as originally thought, but its main conjugates p-cresyl sulphate (p-CS) and p-cresyl glucuronide (p-CG) accumulated in uraemia. As it crosses the intestinal mucosa and liver, p-cresol is detoxified by sulphatation and glucuronidation generating p-CS and p-CG, respectively [1–3]. The toxicity of p-CS has been largely evidenced through several clinical studies showing that serum levels of free p-CS are strongly associated with cardiovascular risk and mortality [4, 5] and experimental studies which demonstrated that p-CS induces oxidative stress, endothelial dysfunction and proximal tubular injury [6–9]. In a previous study, we demonstrated that p-CS induces insulin resistance in vivo in mice as well as in vitro [10], which could partly explain the relationship between insulin resistance and cardiovascular risk in CKD in clinical outcome studies [11]. Surprisingly, p-CS has been extensively studied while its counterpart p-CG has received only limited attention.
Microbial metabolites originating from colonic protein fermentation could be different between humans and rodents due to large differences in microbiota and xenobiotic metabolism. Such differences should be taken into consideration when transposing the results from experimental studies on rodents to humans. Several human studies have highlight that total and free serum levels of p-CS are substantially higher than those of p-CG [12–14]. Reports of coupled p-CS and p-CG concentrations in CKD patients are scarce and the limited existing reports show large discrepancies. Poesen et al. reported total p-CS levels to be roughly 190-fold higher than those of p-CG [3], while Liabeuf et al. found that p-CS concentration was approximately 7- to 9-fold higher [12]. Itoh et al. [14] reported a 4.6- to 23.6-fold higher concentration of pCS versus pCG in haemodialysis patients. It is clear that among CKD patients on dialysis as well as before initiation of renal suppletion therapy, the ratio of p-CS and p-CG concentrations exhibits large variations and the reasons still remain unclear. In any case, p-CS is nowadays widely accepted as the main metabolite of p-cresol in humans. p-CG and p-CS exhibit large differences in their binding to proteins, with p-CS being highly bound (>90%) and p-CG only weakly bound (<10%). However, since most of p-CS is bound while most p-CG is free, their absolute free concentrations are in the same range [12]. In contrast to these human data, after intra-peritoneal injection of p-cresol into laboratory rats, 64% of the initially injected p-cresol was excreted as a glucuronide conjugate [15]. Mice are commonly used as an animal model of CKD because they are easy to handle and transgenic strains are widely available. There are, however, no data on the relative proportion of p-CG and p-CS in mice. Furthermore, the biological effect of chronic administration of p-cresol and its metabolite p-CG remains to be explored. The fact that p-CG only recently became commercially available could account for this lack of data in the literature.
A better understanding of the biological impact of p-CG is needed, because while being a minor metabolite of p-cresol in humans, both total and free p-CG levels were positively associated with overall and cardiovascular mortality independently of common well-known predictors of survival [3, 12]. To the best of our knowledge, there is however limited data on the putative biological effects of p-CG. The experimental studies investigating the biological activity and toxicity of p-CG produced contradictory results. On one hand, p-CG failed to enhance apoptosis, and to promote a pro-inflammatory phenotype in renal tubular cells [16]. p-CG per se did not increase leucocyte oxidative burst activity, suggesting that p-CG might be a bio-inactive compound [13]. On the other hand, a synergistic effect of p-CS and p-CG in activation of leucocytes was observed in vitro [13, 17] but not in vivo [17]. A direct effect of p-CG was described on mitochondrial succinate dehydrogenase activity [18] or efflux transporters in renal tubular cells [19]. Recently, a relative shift from sulphate to glucuronide conjugation of p-cresol was observed in advanced CKD patients, suggesting a blunted sulphotransferase activity in CKD that was independently associated with mortality and cardiovascular diseases [3].
The purpose of this study was (i) to evaluate the relative importance of the main metabolites of p-cresol in mice with normal kidney function and with kidney failure, and (ii) to assess in vitro and in vivo the impact on insulin resistance of p-CS and p-CG alone or of a combination of both.
MATERIALS AND METHODS
Chemicals and antibodies
All chemicals and reagents were obtained from Sigma Aldrich (Saint-Quentin Fallavier, France) when no other origin is specified. Anti-Phospho-Akt 1/2/3 and anti-Akt 1/2/3 antibodies were purchased from Santa Cruz Biotechnology (Heidelberg, Germany). Anti-rabbit Immunoglobulin G(IgG) antibodies were from BioRad (Marnes-la-Coquette, France).
p-Cresol, p-CS and p-CG
p-Cresol was obtained from Sigma Aldrich. The potassium salts of p-CS and p-CG were synthesized according to Feigenbaum and Neuberg [20] and Desai and Blackwell [21], respectively. In vitro the concentration of p-cresol was 40 mg/L and the concentration of p-CS was 40 mg/L, which is the concentration found in humans with end-stage renal disease (ESRD) [22]. We chose a p-CG concentration of 6 mg/L according to Itoh et al. [14] and a 10-fold higher concentration (i.e. 60 mg/L). All experiments were conducted according to the standardized approach for in vitro research in uraemia [12, 13, 23].
Animals
Animal experiments were performed under the authorization n°69-266-0501 (INSA-Lyon, Direction Départementale de la Protection des Populations - Services Vétérinaires - DDPP-SV), according to the guidelines laid down by European Union Council Directive 2010/63UE. CD1 Swiss mice were purchased from Janvier-Labs (Le Genest-Saint-Isle, France) and housed in an air-conditioned room with a controlled environment of 21 ± 0.5°C and 60–70% humidity, under a 12 h light/dark cycle (light on from 07:00 h to 19:00 h) with free access to food and water. CKD was induced by 5/6 nephrectomy with a two-step surgical procedure as previously described [10, 24] and mice were used after 3–4 weeks of uraemia.
p-Cresol, p-CS and p-CG treatment
Mice were randomly assigned to receive twice daily intraperitoneal (i.p.) injections of p-cresol (10 mg/kg), p-CG (10 mg/kg), p-CS (10 mg/kg) or vehicle for control mice, for 4 weeks. Food intake and body weight (BW) were measured twice weekly.
p-CG pharmacokinetic study
Mice were injected with p-CG (10 mg/kg i.p.) and three mice were sacrificed at each time point (i.e. 10, 30, 60 and 240 min after injection) to collect blood. Mice were anaesthetized with isoflurane and blood was collected by heart puncture. Blood was maintained at 4°C overnight to clot to obtain serum. Serum samples were collected, snap frozen in liquid nitrogen and stored at −80°C until p-CG assay.
Insulin tolerance test
After 3 weeks of treatment with p-cresol or p-CG, insulin tolerance tests (ITTs) were performed. After an overnight fast, animals were injected i.p. with 0.5 UI/kg BW of recombinant human insulin (Actrapid® Novo Nordisk, La Défense, France). Blood glucose was measured on a drop of blood sampled from the terminal portion of the tail using a glucometer (Accu-Check Performa, Roche, Meylan, France), before and 15, 30, 60 and 120 min after insulin injection.
Sacrifice and tissue dissection
After 4 weeks of treatment with p-cresol or p-CG, mice were sacrificed. Animals were anaesthetized with sodium pentobarbital (100 mg/kg i.p.). Blood (750 µL) was collected through cardiac puncture in a heparinized syringe. Blood was immediately centrifuged for 2 min at 3500 g to obtain plasma (plasma biochemistry) or was maintained at 4°C overnight to clot to obtain serum (for p-CS and p-CG assay). Plasma and serum samples were snap frozen in liquid nitrogen and stored at −80°C until analysis. Liver, heart, kidneys, gastrocnemius muscle and epididymal, retroperitoneal and subcutaneous white adipose tissues (WAT) were dissected out, weighed, snap frozen in liquid nitrogen and stored at −80°C.
Biochemical measurements
The concentrations of total and free p-CS and p-CG were quantified in serum by reverse-high performance liquid chromatography (HPLC) coupled to a fluorescence detector as previously described [13]. The limit of detection of the HPLC-fluorescence method used for measurement of serum p-CS and p-CG was 0.08 mg/L. The plasma concentration of total cholesterol and triglycerides were determined using enzymatic kits from bioMérieux (Marcy l’Etoile, France). Muscle and liver lipids were extracted using chloroform–methanol (2:1, v/v) according to Folch et al. [25] and total lipid content was estimated gravimetrically.
Cellularity study: measurement of adipocyte size and number
Preparation of adipose tissue for determination of cell size was performed essentially as previously described [26, 27]. The mean fat cell volume was calculated from cell radius and the fat cell number was calculated by dividing the tissue weight by the mean adipocyte weight.
Cell cultures, viability test and insulin stimulation
C2C12 myoblasts (CRL-1772, American Type Culture Collection (ATCC)) were cultured in Dulbecco’s modified Eagle’s medium supplemented with 10% (v/v) heat-inactivated fetal bovine serum as previously described [10]. To test the effects of p-CS and p-CG on differentiated C2C12 myotubes, cell toxicity was determined using MTT assay (Roche). To assess the insulin sensibility of C2C12 myotubes, cells were stimulated with 100 nM insulin (Actrapid®, 100 nM) for 20 min in the presence or absence of p-cresol, p-CS and/or p-CG with bovine serum albumin (BSA) (35 g/L) for 30 min.
Glucose uptake assay
C2C12 cells grown in 12-well plates were treated in the presence or absence with of p-cresol, p-CS and/or p-CG for 30 min and glucose uptake was measured using tritiated 2-deoxy-d-glucose as previously described [10, 28].
Insulin signalling, gel electrophoresis and western blotting
To explore insulin signalling in C2C12 myotubes, cells were stimulated with 100 nM insulin (Actrapid®, 100 nM) for 20 min in the presence or absence of p-cresol, p-CS and/or p-CG with BSA (35 g/L) for 30 min. C2C12 cells were scraped on ice in standard lysis buffer (20 mM Tris, 138 mM NaCl, 2.7 mM KCl, 1 mM MgCl2, 5% glycerol, 1%, NP40 and supplemented extemporaneously with 5 mM ethylenediaminetetraacetic acid, 1 mM Na3VO4, 20 mM NaF, 1 mM dithiothreitol and a protease inhibitor cocktail) and centrifuged (13 000 g, 15 min, 4°C). Protein concentration of cell supernatants was determined with Bradford protein assay. Forty micrograms of protein lysate were separated in SDS-polyacrylamide gels (10%) and transferred onto a nitrocellulose membrane. After transfer, the membranes were blocked with 5% (w/v) BSA in tris-buffered saline (TBS)-Tween for 2 h. Blots were probed overnight with specific primary antibodies at 4°C followed by 1-h incubation at room temperature with secondary antibodies. Protein bands were detected with the enhanced chemiluminescence substrate kit (Supersignal® West Pico, Perbio, Brebières, France).
Statistical analysis
Data were expressed as means ± standard error of the mean (SEM). All data were analysed using GraphPad Prism v5.0 software (GraphPad software, La Jolla, CA, USA). Simple comparisons were performed using Student’s t-test. Multiple comparisons were performed using ANOVA followed when appropriate by post hoc Fisher Protected Least Significant Difference (PLSD) tests. Differences were considered significant at the P < 0.05 level.
RESULTS
Serum concentration of p-CS and p-CG in mice
The concentrations of p-CS and p-CG in control mice were 0.20 and 0.23 mg/L, respectively. Nephrectomized mice exhibited a very significant increase of both compounds (1.53 and 1.37 mg/L for p-CS and p-CG, respectively). Of note in control and in CKD mice, the free p-CS concentration was below the limit of quantification. The proportional concentrations (%) of p-CG and p-CG in different situations are illustrated Figure 1 and Supplementary Table S1. In humans (Figure 1A), the main metabolite of p-cresol is p-CS (87%) while p-CG is only a minor metabolite (13%). In contrast, in control and CKD mice (Figure 1B and C) the proportion of p-CS and p-CG was similar (42% and 52% in control mice and 53% and 47% in CKD mice, respectively).
To reproduce the repartition and levels of metabolites of p-cresol in the uraemic mice, mice were injected with p-cresol, the parent compound of p-CS and p-CG. After injection of p-cresol, we observed an increase of total p-CS and p-CG in the same proportion and concentration (46% of total p-CS and 54% of total p-CG) as observed in CKD mice (Figure 1D and Supplementary Table S1). Mice injected with either p-CS (Figure 1E and Table Supplementary S1) or p-CG (Figure 1F and Supplementary Table S1) exhibited a major increase in the proportion of the administered metabolite with negligible conversion from one form to another. Interestingly, the total concentrations of p-CS and p-CG in control and CKD mice were highly correlated (r2 = 0.924, P < 0.0001) (Figure 2). In good agreement with previous reports in human [12], p-CG was poorly bound (<6%) while p-CS was mainly bound (96%) to plasma proteins in mice (Supplementary Table S1).
Chronic administration of p-cresol induces metabolic disturbances in mice
The results of the ITT after 4 weeks of p-cresol administration are presented in Figure 3A. Insulin administration triggered a significantly (P < 0.001) larger hypoglycaemic response in control mice (−70%) than in p-cresol-treated mice (−51%) (Figure 3A and B). These results are consistent with the decrease in insulin sensitivity observed in CKD mice in which concentrations of p-CS and p-CG reached the same levels [10]. Moreover, in p-cresol-treated mice, plasma cholesterol levels were increased by 27% (P < 0.01) compared with vehicle-treated mice, while, the plasma concentration of triglycerides and glucose were not different (Table 1). These results are consistent with the metabolic profile previously reported for uraemic mice [10, 29].
Treatment . | Vehicle . | p-Cresol . | Change (%) . | P-value . |
---|---|---|---|---|
Food intake, kJ/24 h | 10.3 ± 0.7 | 10.6 ± 0.8 | 3 | 0.792 |
Biometric data | ||||
BW, g | 30 ± 1.0 | 28 ± 1.0 | −7 | 0.098 |
Body length, cm | 9.8 ± 0.1 | 9.7 ± 0.1 | 1 | 0.241 |
Lee index, × 103 | 314 ± 2.0 | 313 ± 2.0 | 0 | 0.709 |
Liver, mg/10 g | 428 ± 21.0 | 453 ± 24.0 | 6 | 0.848 |
Kidney, mg/10 g | 120 ± 5.0 | 127 ± 5.0 | 6 | 0.779 |
Heart, mg/10 g | 46 ± 1.0 | 50 ± 1.0 | 9 | 0.750 |
Gastrocnemius, mg/10 g | 51 ± 2.0 | 52 ± 2.0 | 2 | 0.504 |
Total WAT, mg/10 g | 504 ± 57.0 | 397 ± 71 | −27 | 0.008* |
Plasma metabolites | ||||
Urea, mmol/L | ||||
Triglycerides, mmol/L | 1.3 ± 0.3 | 1.3 ± 0.2 | −3 | 0.87 |
Total cholesterol, mmol/L | 2.9 ± 0.2 | 3.3 ± 0.4 | +27 | 0.01 |
Glucose, mmol/L | 7.6 ± 0.3 | 6.4 ± 2.0 | 0 | 0.09 |
Treatment . | Vehicle . | p-Cresol . | Change (%) . | P-value . |
---|---|---|---|---|
Food intake, kJ/24 h | 10.3 ± 0.7 | 10.6 ± 0.8 | 3 | 0.792 |
Biometric data | ||||
BW, g | 30 ± 1.0 | 28 ± 1.0 | −7 | 0.098 |
Body length, cm | 9.8 ± 0.1 | 9.7 ± 0.1 | 1 | 0.241 |
Lee index, × 103 | 314 ± 2.0 | 313 ± 2.0 | 0 | 0.709 |
Liver, mg/10 g | 428 ± 21.0 | 453 ± 24.0 | 6 | 0.848 |
Kidney, mg/10 g | 120 ± 5.0 | 127 ± 5.0 | 6 | 0.779 |
Heart, mg/10 g | 46 ± 1.0 | 50 ± 1.0 | 9 | 0.750 |
Gastrocnemius, mg/10 g | 51 ± 2.0 | 52 ± 2.0 | 2 | 0.504 |
Total WAT, mg/10 g | 504 ± 57.0 | 397 ± 71 | −27 | 0.008* |
Plasma metabolites | ||||
Urea, mmol/L | ||||
Triglycerides, mmol/L | 1.3 ± 0.3 | 1.3 ± 0.2 | −3 | 0.87 |
Total cholesterol, mmol/L | 2.9 ± 0.2 | 3.3 ± 0.4 | +27 | 0.01 |
Glucose, mmol/L | 7.6 ± 0.3 | 6.4 ± 2.0 | 0 | 0.09 |
Data are mean ± SEM and variation compared with control group.
Differences were considered significant at the P < 0.05 level using Student’s t-test.
n = 10 for each group.
Treatment . | Vehicle . | p-Cresol . | Change (%) . | P-value . |
---|---|---|---|---|
Food intake, kJ/24 h | 10.3 ± 0.7 | 10.6 ± 0.8 | 3 | 0.792 |
Biometric data | ||||
BW, g | 30 ± 1.0 | 28 ± 1.0 | −7 | 0.098 |
Body length, cm | 9.8 ± 0.1 | 9.7 ± 0.1 | 1 | 0.241 |
Lee index, × 103 | 314 ± 2.0 | 313 ± 2.0 | 0 | 0.709 |
Liver, mg/10 g | 428 ± 21.0 | 453 ± 24.0 | 6 | 0.848 |
Kidney, mg/10 g | 120 ± 5.0 | 127 ± 5.0 | 6 | 0.779 |
Heart, mg/10 g | 46 ± 1.0 | 50 ± 1.0 | 9 | 0.750 |
Gastrocnemius, mg/10 g | 51 ± 2.0 | 52 ± 2.0 | 2 | 0.504 |
Total WAT, mg/10 g | 504 ± 57.0 | 397 ± 71 | −27 | 0.008* |
Plasma metabolites | ||||
Urea, mmol/L | ||||
Triglycerides, mmol/L | 1.3 ± 0.3 | 1.3 ± 0.2 | −3 | 0.87 |
Total cholesterol, mmol/L | 2.9 ± 0.2 | 3.3 ± 0.4 | +27 | 0.01 |
Glucose, mmol/L | 7.6 ± 0.3 | 6.4 ± 2.0 | 0 | 0.09 |
Treatment . | Vehicle . | p-Cresol . | Change (%) . | P-value . |
---|---|---|---|---|
Food intake, kJ/24 h | 10.3 ± 0.7 | 10.6 ± 0.8 | 3 | 0.792 |
Biometric data | ||||
BW, g | 30 ± 1.0 | 28 ± 1.0 | −7 | 0.098 |
Body length, cm | 9.8 ± 0.1 | 9.7 ± 0.1 | 1 | 0.241 |
Lee index, × 103 | 314 ± 2.0 | 313 ± 2.0 | 0 | 0.709 |
Liver, mg/10 g | 428 ± 21.0 | 453 ± 24.0 | 6 | 0.848 |
Kidney, mg/10 g | 120 ± 5.0 | 127 ± 5.0 | 6 | 0.779 |
Heart, mg/10 g | 46 ± 1.0 | 50 ± 1.0 | 9 | 0.750 |
Gastrocnemius, mg/10 g | 51 ± 2.0 | 52 ± 2.0 | 2 | 0.504 |
Total WAT, mg/10 g | 504 ± 57.0 | 397 ± 71 | −27 | 0.008* |
Plasma metabolites | ||||
Urea, mmol/L | ||||
Triglycerides, mmol/L | 1.3 ± 0.3 | 1.3 ± 0.2 | −3 | 0.87 |
Total cholesterol, mmol/L | 2.9 ± 0.2 | 3.3 ± 0.4 | +27 | 0.01 |
Glucose, mmol/L | 7.6 ± 0.3 | 6.4 ± 2.0 | 0 | 0.09 |
Data are mean ± SEM and variation compared with control group.
Differences were considered significant at the P < 0.05 level using Student’s t-test.
n = 10 for each group.
Biometric data for each group are also shown in Table 1. No significant difference in food intake, BW and lean body mass were noticed between p-cresol mice and control mice. However, p-cresol mice exhibited a significant decrease in WAT accretion (−27%, P < 0.008) (Figure 3C). p-Cresol treatment significantly decreased WAT mass in the intra-abdominal fat pads, e.g. the epididymal (−51%, P < 0.05) and the retroperitoneal (−64%, P < 0.05) fat pads, as well as in the subcutaneous inguinal fat pad (−38%, P < 0.05) (Figure 3C). As previously described in CKD mice [10, 24, 30, 31], ectopic lipid redistribution was observed in skeletal muscle (+47%, P < 0.05) and liver (+20%, P < 0.05) compared with control (Figure 3D), suggesting lipotoxicity as a putative cause for insulin resistance [32]. To analyse whether changes in WAT accretion resulted from a reduction in the number of adipocytes or a decreased triglyceride accumulation in adipocytes, we performed a cytological analysis of epididymal WAT pads. The cellular characteristics of epididymal WAT in control and p-cresol-treated mice are described in Table 2. Mean adipocyte diameter was reduced by 10% in mice injected with p-cresol (P < 0.05) resulting in a 30% decrease in adipose cell weight (P < 0.05). In contrast, the total number of adipocytes per epidymal fat pad was not significantly altered, suggesting that decreased fat mass resulted from a hypotrophia (i.e. a decreased cell size) rather than hypoplasia (i.e. a decreased number of cells).
Treatment . | Control . | p-Cresol . | Change (%) . | P-value . |
---|---|---|---|---|
n | 8 | 9 | ||
eWAT, mg | 535 ± 113 | 265 ± 26 | −50 | 0.032* |
Cell diameter, μm | 52 ± 2.0 | 47 ± 1.0 | −10 | 0.035* |
Cell weight, ng | 93 ± 10.0 | 65 ± 6.0 | −30 | 0.024* |
No. of cells, × 106 | 5.1 ± 0.6 | 4.1 ± 0.8 | −21 | 0.180 |
Treatment . | Control . | p-Cresol . | Change (%) . | P-value . |
---|---|---|---|---|
n | 8 | 9 | ||
eWAT, mg | 535 ± 113 | 265 ± 26 | −50 | 0.032* |
Cell diameter, μm | 52 ± 2.0 | 47 ± 1.0 | −10 | 0.035* |
Cell weight, ng | 93 ± 10.0 | 65 ± 6.0 | −30 | 0.024* |
No. of cells, × 106 | 5.1 ± 0.6 | 4.1 ± 0.8 | −21 | 0.180 |
Data are mean ± SEM and variation compared with control group.
Differences were considered significant at the P < 0.05 level using Student’s t-test.
Treatment . | Control . | p-Cresol . | Change (%) . | P-value . |
---|---|---|---|---|
n | 8 | 9 | ||
eWAT, mg | 535 ± 113 | 265 ± 26 | −50 | 0.032* |
Cell diameter, μm | 52 ± 2.0 | 47 ± 1.0 | −10 | 0.035* |
Cell weight, ng | 93 ± 10.0 | 65 ± 6.0 | −30 | 0.024* |
No. of cells, × 106 | 5.1 ± 0.6 | 4.1 ± 0.8 | −21 | 0.180 |
Treatment . | Control . | p-Cresol . | Change (%) . | P-value . |
---|---|---|---|---|
n | 8 | 9 | ||
eWAT, mg | 535 ± 113 | 265 ± 26 | −50 | 0.032* |
Cell diameter, μm | 52 ± 2.0 | 47 ± 1.0 | −10 | 0.035* |
Cell weight, ng | 93 ± 10.0 | 65 ± 6.0 | −30 | 0.024* |
No. of cells, × 106 | 5.1 ± 0.6 | 4.1 ± 0.8 | −21 | 0.180 |
Data are mean ± SEM and variation compared with control group.
Differences were considered significant at the P < 0.05 level using Student’s t-test.
p-CG does not induce insulin resistance
We previously reported that p-CS promotes insulin resistance in CKD [10, 33] by interfering with insulin signalling pathways. To gain insights into the biological impact of p-CG, we incubated C2C12 muscle cells with p-CS or p-CG at the concentration found in ESRD patients or p-cresol as their precursor. No significant difference of cell viability was observed between control and cells incubated with p-cresol, p-CS or p-CG at relevant concentrations. A minor cytotoxic effect was only noticed for p-cresol concentration higher than 80 mg/L (P < 0.05). Treatment of cells with both p-CS (40 mg/L) and p-CG (6 mg/L) together, i.e. at concentrations found in ESRD patients, however, significantly decreased cell viability (−8%, P < 0.05) (Supplementary Table S2). The effect of p-cresol, p-CG, p-CS and the combination of p-CS and p-CG on glucose uptake and PKB/Akt phosphorylation is illustrated in Figure 4. None of the salt control solutions (KCl or K2SO4) triggered a significant difference versus saline (data not shown). Therefore, only saline control data are presented. p-CG (61 mg/L) by itself had no impact on the insulin-stimulated glucose uptake in C2C12 cells. In contrast, p-cresol and p-CS (40 mg/L) abolished the insulin-stimulated glucose uptake without affecting the basal glucose uptake (P < 0.05) (Figure 4A). In good agreement, insulin-induced serine-phosphorylation of PKB/Akt was totally inhibited after p-cresol and p-CS treatment compared with control, while p-CG had no effect on the insulin signalling pathway (Figure 4B and C). The combination of p-CS and p-CG also inhibited insulin-induced serine-phosphorylation of PKB/Akt, but to an extent that was not different from the one reached by p-CS alone. Of note, the combination of both compounds did not exhibit any synergistic effect compared with p-CS alone.
To investigate the effect of p-CG on metabolism, mice were given 10 mg/kg twice daily for 4 weeks. Intraperitoneal injections of p-CG in mice with normal renal function induced a transient increase in serum p-CG concentration. The concentration of serum total p-CG reached 18 mg/L 30 min after the injection and its concentration dropped to 2.3 mg/L in 1 h (Figure 5A). Mice exhibited a free p-CG serum level of 16 mg/L at 30 min and of 2.2 mg/dL after 1 h (Figure 5B). In good agreement with in vitro studies, chronic treatment with p-CG failed to impair insulin sensitivity in mice. The decrease in blood glucose in response to insulin was similar in control and p-CG mice (Figure 6A and B). Fed glycaemia, total cholesterol and triglycerides concentrations were similar in control and p-CG, while fasting glycaemia was slightly increased in p-CG mice (+17%, P < 0.05) (Table 3). The food intake, BW and lean mass were not different between groups (Table 3). p-CG mice exhibited only a minor change in WAT accretion (Table 3). Indeed, as shown in Figure 6C, changes were restricted to subcutaneous fat pad while total fat mass remained unaffected. p-CG had no significant effect on renal function (Table 3).
Treatment . | Control . | p-CG . | Change (%) . | P-value . |
---|---|---|---|---|
Food intake, kJ/24 h | 8.6 ± 0.8 | 8.7 ± 0.8 | 1 | 0.929 |
Biometric data | ||||
BW, g | 34 ± 1.0 | 33 ± 1 | −3 | 0.755 |
Body length, cm | 10.6 ± 0.1 | 10.7 ± 0.1 | 6 | 0.692 |
Lee index, × 103 | 304 ± 3.0 | 302 ± 2 | −1 | 0.439 |
Liver, mg/10 g | 453 ± 21 | 459 ± 24 | 1 | 0.797 |
Kidney, mg/10 g | 114 ± 3.0 | 107 ± 5 | −6 | 0.172 |
Heart, mg/10 g | 42 ± 1.0 | 45 ± 1 | 7 | 0.051 |
Gastrocnemius, mg/10 g | 43 ± 2.0 | 48 ± 2 | 11 | 0.227 |
Total WAT, mg/10 g | 702 ± 87 | 504 ± 57 | −28 | 0.075 |
Plasma metabolites | ||||
Fasting glycaemia, mmol/L | 3.54 ± 0.17 | 4.16 ± 0.21 | +17 | 0.036* |
Fed glycaemia, mmol/L | 6.99 ± 0.31 | 6.77 ± 0.28 | −3 | 0.601 |
Triglycerides, mmol/L | 1.6 ± 0.23 | 1.15 ± 0.14 | −30 | 0.1090 |
Total cholesterol, mmol/L | 3.27 ± 0.38 | 2.99 ± 0.17 | −9 | 0.501 |
Index of renal function | ||||
Urea, mmol/L | 42 ± 1.0 | 45 ± 1 | 7 | 0.051 |
24 h diuresis, mL/24 h | 0.7 ± 0.1 | 0.5 ± 0.1 | −29 | 0.319 |
Proteinuria, mg/24 h | 0.21 ± 0.05 | 0.19 ± 0.09 | −5 | 0.898 |
Treatment . | Control . | p-CG . | Change (%) . | P-value . |
---|---|---|---|---|
Food intake, kJ/24 h | 8.6 ± 0.8 | 8.7 ± 0.8 | 1 | 0.929 |
Biometric data | ||||
BW, g | 34 ± 1.0 | 33 ± 1 | −3 | 0.755 |
Body length, cm | 10.6 ± 0.1 | 10.7 ± 0.1 | 6 | 0.692 |
Lee index, × 103 | 304 ± 3.0 | 302 ± 2 | −1 | 0.439 |
Liver, mg/10 g | 453 ± 21 | 459 ± 24 | 1 | 0.797 |
Kidney, mg/10 g | 114 ± 3.0 | 107 ± 5 | −6 | 0.172 |
Heart, mg/10 g | 42 ± 1.0 | 45 ± 1 | 7 | 0.051 |
Gastrocnemius, mg/10 g | 43 ± 2.0 | 48 ± 2 | 11 | 0.227 |
Total WAT, mg/10 g | 702 ± 87 | 504 ± 57 | −28 | 0.075 |
Plasma metabolites | ||||
Fasting glycaemia, mmol/L | 3.54 ± 0.17 | 4.16 ± 0.21 | +17 | 0.036* |
Fed glycaemia, mmol/L | 6.99 ± 0.31 | 6.77 ± 0.28 | −3 | 0.601 |
Triglycerides, mmol/L | 1.6 ± 0.23 | 1.15 ± 0.14 | −30 | 0.1090 |
Total cholesterol, mmol/L | 3.27 ± 0.38 | 2.99 ± 0.17 | −9 | 0.501 |
Index of renal function | ||||
Urea, mmol/L | 42 ± 1.0 | 45 ± 1 | 7 | 0.051 |
24 h diuresis, mL/24 h | 0.7 ± 0.1 | 0.5 ± 0.1 | −29 | 0.319 |
Proteinuria, mg/24 h | 0.21 ± 0.05 | 0.19 ± 0.09 | −5 | 0.898 |
Data are mean ± SEM and variation compared to control group.
Differences were considered significant at the P < 0.05 level using Student’s t-test. n = 9 for each group.
Treatment . | Control . | p-CG . | Change (%) . | P-value . |
---|---|---|---|---|
Food intake, kJ/24 h | 8.6 ± 0.8 | 8.7 ± 0.8 | 1 | 0.929 |
Biometric data | ||||
BW, g | 34 ± 1.0 | 33 ± 1 | −3 | 0.755 |
Body length, cm | 10.6 ± 0.1 | 10.7 ± 0.1 | 6 | 0.692 |
Lee index, × 103 | 304 ± 3.0 | 302 ± 2 | −1 | 0.439 |
Liver, mg/10 g | 453 ± 21 | 459 ± 24 | 1 | 0.797 |
Kidney, mg/10 g | 114 ± 3.0 | 107 ± 5 | −6 | 0.172 |
Heart, mg/10 g | 42 ± 1.0 | 45 ± 1 | 7 | 0.051 |
Gastrocnemius, mg/10 g | 43 ± 2.0 | 48 ± 2 | 11 | 0.227 |
Total WAT, mg/10 g | 702 ± 87 | 504 ± 57 | −28 | 0.075 |
Plasma metabolites | ||||
Fasting glycaemia, mmol/L | 3.54 ± 0.17 | 4.16 ± 0.21 | +17 | 0.036* |
Fed glycaemia, mmol/L | 6.99 ± 0.31 | 6.77 ± 0.28 | −3 | 0.601 |
Triglycerides, mmol/L | 1.6 ± 0.23 | 1.15 ± 0.14 | −30 | 0.1090 |
Total cholesterol, mmol/L | 3.27 ± 0.38 | 2.99 ± 0.17 | −9 | 0.501 |
Index of renal function | ||||
Urea, mmol/L | 42 ± 1.0 | 45 ± 1 | 7 | 0.051 |
24 h diuresis, mL/24 h | 0.7 ± 0.1 | 0.5 ± 0.1 | −29 | 0.319 |
Proteinuria, mg/24 h | 0.21 ± 0.05 | 0.19 ± 0.09 | −5 | 0.898 |
Treatment . | Control . | p-CG . | Change (%) . | P-value . |
---|---|---|---|---|
Food intake, kJ/24 h | 8.6 ± 0.8 | 8.7 ± 0.8 | 1 | 0.929 |
Biometric data | ||||
BW, g | 34 ± 1.0 | 33 ± 1 | −3 | 0.755 |
Body length, cm | 10.6 ± 0.1 | 10.7 ± 0.1 | 6 | 0.692 |
Lee index, × 103 | 304 ± 3.0 | 302 ± 2 | −1 | 0.439 |
Liver, mg/10 g | 453 ± 21 | 459 ± 24 | 1 | 0.797 |
Kidney, mg/10 g | 114 ± 3.0 | 107 ± 5 | −6 | 0.172 |
Heart, mg/10 g | 42 ± 1.0 | 45 ± 1 | 7 | 0.051 |
Gastrocnemius, mg/10 g | 43 ± 2.0 | 48 ± 2 | 11 | 0.227 |
Total WAT, mg/10 g | 702 ± 87 | 504 ± 57 | −28 | 0.075 |
Plasma metabolites | ||||
Fasting glycaemia, mmol/L | 3.54 ± 0.17 | 4.16 ± 0.21 | +17 | 0.036* |
Fed glycaemia, mmol/L | 6.99 ± 0.31 | 6.77 ± 0.28 | −3 | 0.601 |
Triglycerides, mmol/L | 1.6 ± 0.23 | 1.15 ± 0.14 | −30 | 0.1090 |
Total cholesterol, mmol/L | 3.27 ± 0.38 | 2.99 ± 0.17 | −9 | 0.501 |
Index of renal function | ||||
Urea, mmol/L | 42 ± 1.0 | 45 ± 1 | 7 | 0.051 |
24 h diuresis, mL/24 h | 0.7 ± 0.1 | 0.5 ± 0.1 | −29 | 0.319 |
Proteinuria, mg/24 h | 0.21 ± 0.05 | 0.19 ± 0.09 | −5 | 0.898 |
Data are mean ± SEM and variation compared to control group.
Differences were considered significant at the P < 0.05 level using Student’s t-test. n = 9 for each group.
DISCUSSION
In this study, we investigated the influence of uraemic toxins derived from phenol metabolism on the insulin resistance associated with CKD. The key findings are: (i) substantial differences are observed between human and mouse in the handling of p-cresol, and (ii) although p-CG concentration in CKD patients was associated with mortality, p-CG alone or in combination with p-CS did not promote insulin resistance, in contrast to p-CS.
First, we demonstrated that p-cresol in mice is metabolized into at least two metabolites, p-CG and p-CS, confirming a previous suggestion by Lesaffer et al. based on data obtained in rats [15, 34]. However, in contrast to humans, in whom p-CS is the main metabolite [1–3], p-CG and p-CS are produced in equal amounts in rodents (i.e. roughly 50/50). Thus, injection of p-cresol into mice with normal renal function mimicked the serum concentration of p-CS and p-CG observed in uraemic animals. To specifically determine the role of p-CG and p-CS, the most direct strategy is, therefore, to administer p-CS or p-CG as such (Figure 1 and Supplementary Table S1). In agreement with previous reports [12–14], we confirm that the conjugates of p-cresol showed a marked difference in percentage binding to protein, with a high binding for p-CS (>95%) and low binding for p-CG (≈5%) (Table 1). In humans, although most of p-CS is bound while p-CG, to a large extent, is free, their absolute free concentrations are in the same range. In humans, the degree of protein binding of p-CG seems to vary among individuals. This was also observed in spiking experiments performed by Poesen et al. [3] where the percentage of binding of p-CG varied from 25% to 12% in healthy versus uraemic serum. As demonstrated in the Supplementary Table S1 in mice, protein binding of p-CG varied from 5.6–10.3% to 23.0%, with the lowest degree of protein binding at the highest p-CG concentration. The contributing factors to this variation are, however, still not clear. A diminished protein binding associated with the decline in renal function have already been reported for p-CS [35]. This could result from the hypoalbuminaemia, chemical modifications or conformational changes of albumin, and/or competitive binding with increasing levels of uraemic retention solutes. Considering that probably only the free fraction is biologically active, determining the role of the free p-CG concentrations as observed in vivo seems crucial.
Secondly, we investigated the role of p-CG on the insulin signalling pathway in vitro as well as in vivo and we are not able to evidence that circulating p-CG at levels as observed in uraemia alters glucose homeostasis. Insulin resistance is a common feature of CKD [36] and is an independent risk factor for cardiovascular morbidity and mortality in these patients [11]. However, the underlying causes of the insulin resistance in CKD remain unclear [29]. In recent years, the specific role of individual uraemic toxins such as urea [37], asymmetric dimethylarginine [38], pseudouridine [39], uric acid [40] and p-CS [10] in the pathogenesis of insulin resistance has been demonstrated. However, in contrast to what was reported in mice undergoing chronic injections of p-CS [10] or p-cresol (Figure 3A), we failed to induce insulin resistance in p-CG‐treated mice (Figure 6A). We only observed a minor increase in fasting glucose level, reflecting still an insubstantial decrease in insulin sensitivity or defect in glucose disposition (Table 3). Of importance, the free and total p-CG concentrations after administration (Supplementary Table S1) were in the range of those observed in ESRD patients [12, 13] and uraemic mice (1.4 versus 2.3 mg/L in CKD and pCG-treated mice, respectively).
The association between ectopic lipid accumulation and insulin resistance has been described in CKD animal models [24, 30–32]. Lipoatrophy and ectopic fat distribution associated with insulin resistance was observed in mice treated with p-cresol (Figure 3B–D) or with p-CS [10], while insignificant changes were noticed in p-CG-treated mice (Figure 6B). These results were confirmed by in vitro data, where the treatment of C2C12 muscle cells with p-CG at concentrations found in ESRD, failed to disrupt insulin signalling (Figure 4). Meert etal. demonstrated in vitro that p-CG alone had no impact on leucocyte activation, but that the addition of p-CG to p-CS exhibited a synergistic leucocyte-activating effect [13]. We incubated C2C12 muscle cell with both p-CG and p-CS, but changes were restricted to a minor decrease on viability while we failed to observe any synergistic effect on inhibition of the insulin pathway. Since p-CG by itself does not interfere with glucose metabolism, the metabolic effect of p-cresol treatment in mice should result from the sole action of p-CS produced from p-cresol metabolism.
Finally, we found in this mouse model a strong correlation between the total concentrations of p-CG and p-CS (Figure 2) as reported in clinical studies [3, 12]. p-Cresol was metabolized into p-CS and p-CG, and hence p-CG concentration is not independent of p-CS concentration. Numerous studies have demonstrated a correlation between mortality and the concentration of p-CS but recent studies reported that also free and total p-CG were significantly associated with mortality [3, 12]. However, as there is only limited evidence that p-CG produces toxic biological effects, although associated with mortality [12], p-CG could be considered as an indirect indicator of the concentration of p-CS.
Sulphatation and glucuronidation are common phase II drug metabolism mechanisms used by the body to facilitate the removal of hydrophobic endogenous compounds. A difference of phase II metabolic clearance between humans and mice has been reported previously, with glucuronides being formed more prominently than sulpho-conjugates in rodents [41]. The negative impact of uraemic medium on sulphotransferase activity, as suggested by Poesen et al. [3], needs to be further explored. Total renal clearance of p-CS was substantially lower than total renal clearance of p-CG. Such differences in renal handling of these two metabolites could also account for the differences observed in toxicity in vivo.
A limitation of the present study was that p-cresol was administered intraperitoneally and we cannot rule out a change in metabolism via the intestinal barrier. Indeed, the intestinal barrier is believed to play a dominant role in phase II conjugation during the first-pass metabolism of phenols [42] and allocation of metabolites after injection of p-cresol may be different. However, the relative proportion of p-CG and p-CS was very similar to that of uraemic or control mice, so we could reasonably expect that bypassing the intestinal barrier in this model has a limited impact. The present study was strictly restricted to deciphering the metabolic effects of p-cresol derivatives and does not rule out the putative impact of these compounds on other biological systems as have been published by Meert et al. and Pletinck et al. [13, 17].
In conclusion, in contrast to p-CS [10], we have found no implication of p-CG in the metabolic complications associated with CKD. However, special care should be taken to transpose mouse results into humans as our data demonstrate obvious inter-species differences.
SUPPLEMENTARY DATA
Supplementary data are available online at http://ndt.oxfordjournals.org.
AUTHORS' CONTRIBUTIONS
L.K. and C.O.S. conceived and designed the study. L.K., P.M.A., M.L.C. and C.O.S. performed experiments. L.K. and C.O.S. performed statistical analysis, interpreted results, wrote and revised the manuscript. S.C. performed the chemical synthesis of p-CS and p-CG. G.G. and R.V. performed and analysed the HPLC measurement of plasma p-CS and p-CG. D.F. edited and revised the manuscript. C.O.S. is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
FUNDING
L.K. held a fellowship from ‘Fondation pour la Recherche Médicale’ and P.M.A. held a grant from ‘Société Française de Néphrologie’. M.L.C. held a grant from the French ‘Ministère de l’Enseignement Supérieur et de la Recherche’. This work was supported by Institut National de la Santé et de la Recherche Médicale (INSERM), Institut National des Sciences Appliquées de Lyon (INSA-Lyon) and Fédération Nationale d’Aide aux Insuffisants Rénaux (FNAIR).
CONFLICT OF INTEREST STATEMENT
None declared.
Comments