Warning: the unfortunate end of p-cresol as a uraemic toxin

Introduction

The cresols are protein-bound uraemic retention solutes that have become a frequent subject of both in vivo and in vitro research. These investigations essentially focused on the mother compound p-cresol. Five years ago, two research groups independently demonstrated that not p-cresol but its conjugates, and among these especially p-cresylsulfate, predominate in the body [1, 2]. Nevertheless, even nowadays, studies are performed whereby only the effects of p-cresol are evaluated, in spite of the likelihood that such an approach is pathophysiologically irrelevant.

The present publication aims to explain the origins for this persisting misunderstanding, warns against emanating conceptual error and outlines a correct approach for the future.

Protein-bound uraemic retention solutes and their relation to other uraemic toxins

Once a certain degree of renal dysfunction is reached, it is often associated with the development of complications, affecting both quality of life and survival [3–6]. This process is paralleled by the retention of a host of compounds (uraemic solutes) [7, 8] to which a key role in this functional deterioration has been attributed.

Uraemic retention products conventionally are classified into three physicochemical groups [7]: next to the small water-soluble compounds and the larger middle molecules (mostly peptidic compounds with molecular weight >500Da) there is the third group, to which the cresols belong, the protein-bound compounds. They have many biological and biochemical (toxic) actions [9], while the height of their concentrations has recently been associated with worsening outcomes [10–12]. Their removal by dialysis is not affected by dialyser pore size [13] and only marginally by convection [14].

The large group of protein-bound uraemic solutes should not be considered as a homogenous entity since pathophysiology, degree and strength of protein binding, protein-binding sites, retention pattern, removal during dialysis and even the nature of the binding protein(s) are unlikely to be the same for all of them. Thus, each subgroup should be considered as a separate entity.

Phenols: a specific subgroup of protein-bound solutes

The phenols, as derived from the amino acid tyrosine, are chemically characterized by a cyclic ring structure containing six carbon atoms to which at least one hydroxyl group is linked. p-Cresol is a phenol derived from the tyrosine carrying one methyl group in the para position (Figure 1). Both phenol and p-cresol are precursors of uraemic retention compounds and are generated by the intestinal flora [15, 16] although environmental factors, herbal substances and alternative or traditional medicines might be additional sources [17].

Fig. 1.

Chemical pathways of generation of p-cresol, p-cresylglucuronide and p-cresylsulfate starting from tyrosine.

Open in new tabDownload slide

The p-cresol/p-cresylsulfate conundrum

Cresol compounds started to be considered more extensively by the uraemic toxin researchers and the nephrological community at large at the end of previous century, mainly due to Japanese pioneering work [18]. From the very beginning, most of the attention went to p-cresol, to which a number of important toxic effects, such as endothelial and immunological dysfunction, were attributed [19–22]. In several in vivo analytical assessments, p-cresol was found in serum samples from normal subjects and uraemic patients [18, 23], with a gradual rise in concentration at progression of renal failure. However, virtually all these determinations were performed after strong acidification for deproteinization. In 2005, Martinez et al. [2] deproteinized without acidification by addition of methanol and observed how virtually no p-cresol could be detected in serum from haemodialysis patients. In contrast, they found substantial amounts of its conjugate, p-cresylsulfate. Independently and simultaneously, de Loor et al. [1] reported similar findings by an indirect approach involving enzymatic degradation. These findings were subsequently corroborated in several other studies applying different deproteinization and analytical methods [10,24–27], all finding increased concentrations of p-cresylsulfate up to mean/median values in the range of 30–45 mg/L [2, 24, 26] and individual values up to a maximum of 80–105 mg/L [10, 26] in dialysed populations (Table 1). De Loor et al. [1] suggested that also another conjugate, p-cresylglucuronide, was present, albeit at markedly lower concentrations than for p-cresylsulfate.

Thus, p-cresol is not the end product of tyrosine since it is further metabolized down the line to sulfate and glucuronide conjugates (Figure 1). The repeated findings of the presence of p-cresol as a uraemic solute in earlier literature are merely attributable to an artifact, induced by hydrolysis of the conjugates p-cresylsulfate and p-cresylglucuronide by strong acidification. Also the differences in measured p-cresol concentrations between research groups applying different methodological approaches [28, 29] can be attributed to this artifact since different degrees of acidification are related to different levels of deconjugation. Indoxyl sulfate, another protein-bound retention solute with similar metabolic roots as p-cresylsulfate, although deriving from tryptophan, is equally degraded by combined acidification and heat but not if it is submitted only to heat, which is an approach frequently applied for indoxyl sulfate determination. Of note, the presence of conjugated phenol esters had been reported already in the literature in the early 1980s [30], but thereafter, this finding was unfortunately neglected.

A change in the conceptual thinking on the pathophysiologic role of the cresol compounds

The finding of the predominant presence of the sulfated conjugate rather than of p-cresol itself necessitated a thorough reshuffling of the pathophysiologic views on the role of the cresols. Since p-cresol is not present in human serum unless at very low concentrations, the theory that this same compound exerted biological and biochemical effects (toxicity) through its presentation to cells and organs via this pathway became untenable. Now, the toxicity of p-cresylsulfate, and potentially also of p-cresylglucuronide, came into focus. The first study demonstrating a toxic effect of a conjugate, by Schepers et al. [31], showed an increased free radical production after exposure of leukocytes to p-cresylsulfate at concentrations usually occurring in uraemia. Of note, these results were entirely opposite to those obtained with the mother compound, p-cresol, which is a strong inhibitor of leukocyte activity [22]. A subsequent study, by Meijers et al. [32], demonstrated an increase of endothelial microparticle release, an indicator of endothelial damage. Both leukocyte activation and endothelial damage contribute to vascular damage [33], one of the main reasons for the morbidity and mortality in chronic kidney disease (CKD) [5].

Clinical studies based on p-cresol: the source of confusion

From 2003 onwards, and up till very recently, several studies associated the concentration of p-cresol in patients with CKD to outcome parameters [34–37]. In all these studies, p-cresol was determined after acidification, so that concentrations in fact are representative of those of the real main retention solute, p-cresylsulfate, as each molecule of the sulfate which is broken down by hydrolysis will generate one molecule of p-cresol. Since the concentration of the glucuronide is proportionally small, whereas the ratio between the glucuronide and the sulfate is unpredictable [1], hydrolysis of the glucuronide likely also contributes to this technically induced generation of p-cresol, but in a less consistent and relatively minor way.

Once it became possible to determine p-cresylsulfate itself, it was demonstrated, more or less as expected, that also this compound was correlated to worse outcomes [10]. Of note, in a recent study in patients with stable angina and an absence of moderate degrees of CKD, p-cresylsulfate concentration correlated with coronary lesions [12], extending the association of this compound with cardiovascular outcomes beyond the scope of uraemia [38].

Although all p-cresol-based studies published after 2005 [35–37] contain extensive comments warning against a misinterpretation by supposing activity of p-cresol itself, this caveat might be overlooked when considering only the sheer data. Therefore, the main message of this monograph is that all p-cresol data in the observational outcome studies mentioned above in no way indicate a toxic effect of p-cresol itself, which should be considered only as a surrogate of the conjugates.

Is there still a place for studying p-cresol?

Although it is clear from the above that the study of toxic effects of p-cresol by itself have become largely obsolete, a few situations could be taken into account where this approach still might be considered.

First of all, p-cresol may remain a topic of interest as a point of comparison to the conjugates to evaluate whether there is a difference in effect, as e.g. demonstrated in the studies on leukocytes by Schepers et al. [31]. However, this only makes sense if the conjugates and the mother compound are evaluated together. Second, the response of any cell in direct contact with p-cresol (e.g. cells of the intestinal wall or cells/organisms living in the intestine like bacteria) might still be worthwhile to investigate. A last option would become only relevant if it is demonstrated that the conjugates are reconverted to p-cresol inside the cell by sulfatase or glucuronidase, so that it is p-cresol again that affects intracellular activities. This option seems rather unlikely, however, at least when considering the completely opposite impact of p-cresol and p-cresylsulfate on leukocytes [31].

Conclusions

Although there is now ample information that concentration of p-cresylsulfate or its surrogate precursor p-cresol is associated with general and cardiovascular outcomes, p-cresylsulfate is the normally occurring principal end product of phenol metabolism. Nevertheless, most of the available information in the literature on biological/biochemical impact concerns p-cresol, which in reality is generated essentially ex vivo by the treatment of blood samples when assessing phenols.

The indirectly determined concentration of p-cresol is representative of the concentration of p-cresylsulfate without being the real culprit for toxicity in the same way as a shadow is representative for the shape and the movements of a person without being responsible of his/her actions.

As a consequence, all studies for toxicity should, with the current state of knowledge, be focused on p-cresylsulfate and p-cresylglucuronide, while studies on the mother compound p-cresol have become largely obsolete.

Conflict of interest statement. None declared.

References