Pre- and Post-Resection Urine Metabolic Profiles of Bladder Cancer Patients: Results of Preliminary Studies on Time Series Metabolomics Analysis

Abstract

The incidence of bladder cancer (BCa) has remained high for many years. Nevertheless, its pathomechanism has not yet been fully understood and is still being studied. Therefore, multiplatform untargeted urinary metabolomics analysis has been performed in order to study differences in the metabolic profiles of urine samples collected at three time points: before transurethral resection of bladder tumor (TURBT), the day after the procedure and two weeks after TURBT. Collected samples were analyzed with the use of high-performance liquid chromatography hyphenated with time-of-flight mass spectrometry detection (HPLC-TOF/MS) and gas chromatography coupled with triple quadrupole mass spectrometry detection (GC-QqQ/MS, in a scan mode). Levels of metabolites selected in our previous study were assessed in order to confirm their potential to differentiate the healthy and diseased samples, regardless of the risk factors and individual characteristics. Hippuric acid, pentanedioic acid and uridine confirmed their potential for sample differentiation. Based on the results of statistical analysis for the paired samples (comparison of metabolic profiles of samples collected before TURBT and two weeks after), a set of metabolites belonging to nucleotide metabolism and methylation processes was also selected. Longitudinal studies proved to be useful for the evaluation of metabolic changes in bladder cancer.

Keywords: TURBT; bladder cancer; metabolic fingerprinting; metabolomics; time series analysis; untargeted metabolomics; urine.

Figure 1

Scheme illustrating sample collection process. Legend: “1” (first time point, first morning urine samples collected on the day of TURBT, before the procedure), “2” (second time point, first morning urine samples collected the day after TURBT), “3” (third time point, first morning urine samples collected 14 days after TURBT).

Figure 2

Exemplary score plots of the Principal Component Analysis models (built with the use of SIMCA-P+ 13.0.3 (Umetrics, Umeå, Sweden)) presenting the clustering of QCs. The clustering assessment was made in order to verify systems stability for analysis using (a) LC-HILIC-MS in positive ionization mode and (b) LC-RP-MS in positive ionization mode. Legend: green circles—“1” (first time point, samples collected before TURBT), blue stars—“2” (second time point, samples collected the day after TURBT), red triangles—“3” (third time point, samples collected 14 days after TURBT), yellow pentagons—QCs. The number of samples: “1” = 10, “2” = 10, “3” = 10, QCs = 7.

Figure 3

Differences in metabolite abundance at three time points for (a) propanoic acid, (b) meso-erythritol, (c) uridine. Signal intensities were normalized by SG. p-values, indicating the statistical significance of the differences in abundance between time points “1” and “3”, for propanoic acid, meso-erythritol and uridine were 0.52, 0.22 and 0.02, respectively. Legend: Line segments, blue and red boxes represent consecutive quartiles. Time points: “1”—samples collected before TURBT; “2”—samples collected the day after TURBT; “3”—samples collected at follow-up visit (14 days after TURBT). The number of samples: “1” = 10, “2” = 10, “3” = 10.

Figure 4

Differences in metabolite abundances at time points “1” (samples collected before TURBT) and “3” (samples collected 14 days after TURBT) for (a) 1,3-dimethyluracil and (b) N1-methyl-2-pyridone-5-carboxamide. Signal intensities were normalized by SG. p-values, indicating statistical significance of the differences in abundance between time points “1” and “3”, for 1,3-dimethyluracil and N1-methyl-2-pyridone-5-carboxamide were 0.02 and 0.01, respectively. Legend: Line segments, blue and red boxes represent consecutive quartiles. Time points: “1”—samples collected before TURBT; “2”—samples collected the day after TURBT; “3”—samples collected at follow-up visit (14 days after TURBT). The number of samples: “1” = 10, “2” = 10, “3” = 10.

Figure 5

Scheme illustrating methylation process, potentially enhanced during BCa, along with related metabolic pathways. Based on [35,36,39].