Towards a better understanding of uraemic molecules

Currently, there is no‐cure or pre‐diagnosis test for patients suffering from Chronic Kidney Disease (CKD). Uremic toxins which normally excreted into the urine by healthy kidneys accumulate in the body due to kidney malfunction causing complications such as higher risk of cardiovascular disease. These complications associated with CKD leading to the high mortality rate among the patients. Although the removal of the uraemic solutes such as urea and creatinine can be achieved by dialysis but it does not affect the survival rate. Therefore, it is suggested that there are some other toxins that cannot be removed by hemodialysis causing the symptoms of CKD. Analysis of serum samples of patients suffering from CKD before and after dialysis compared to samples collected from healthy volunteers can thus give more insight into the effect of uraemic solutes as those remaining after dialysis can lead to the death. In addition, monitoring the changes in metabolites after dialysis can be very helpful to evaluate dialysis membrane efficiency in removal of uraemic toxins. Analyses of uraemic molecules in serum sample of a group of patients suffering from CKD were performed with three analytical methods including capillary electrophoresis‐mass spectrometry (CE‐MS), gas chromatography‐mass spectrometry (GC‐MS) and liquid chromatographymass spectrometry (LC‐MS), and the results compared with a healthy (control) group. Both targeted and non‐targeted (global profiling) studies were aimed. Unlike GC‐MS and LC‐MS which proved to be suitable for the analysis of respectively polar and non‐polar metabolites, CE‐MS practically failed in providing reproducible results to be compared to the other two techniques. Targeted study of some of the known uraemic toxins using GC‐MS was performed employing the available individual standards following suitable derivatisation procedures. On the other hand, higher numbers of known uraemic molecules were identified in LC‐MS upon careful monitoring of MS features such as exact masses of molecular ions and adducts in both positive and negative ion‐modes using available data bases such as METLIN library and Human Metabolome Database. Comparison of the results shows that most of the compounds had significant reduction post dialysis, which also reflects the quality of the dialysis treatment. Also unlike water soluble metabolites, the analysis of protein bound solutes was found less conclusive due to the complications associated with them, for example, in finding a suitable sample preparation to be able to efficiently cleave their bounds with proteins. Global metabolic profiling of the results was also performed employing XCMS platform by visualising the processed data (p‐value < 0.05) as distribution profiles. These profiles generally show significant reduction of detected metabolites after dialysis with metabolites had been more efficiently removed from m/z range of 100 to 500. Accordingly, 50% of metabolites were distributed in m/z range of 250‐550, 25% in the relatively narrower range of 100‐250 and the rest were between m/z of 550 to 950. Also, the majority (90%) of detected metabolites showed relative fold changes less than 70 pre‐dialysis which reduced to about 25 after treatment. Moreover, fold change distributions for 50% of metabolites before dialysis was approximately between 3 and 15 which reduced to less than 5 post dialysis. On the other hand, treatment seems to have insignificant effect on fold changes of about 10% of metabolites. As one application, such information might be beneficial in better understanding of the performance of different dialysis treatments