Changes in the urine volatile metabolome throughout growth of transplanted hepatocarcinoma

Abstract

Trained detection dogs distinguish between urine samples from healthy organisms and organisms with malignant tumors, suggesting that the volatile urine metabolome contains information about tumor progression. The aim of this study was to determine whether the stage of tumor growth affects the chemical differences in the urine of mice and to what extent the "olfactory image of disease" perceived by dogs coincides with the "image of disease" recorded by the mass spectrometer. We used a novel laser ionization mass spectrometry method and propose a mass spectrometric analysis without detailed interpretation of the spectrum of volatile metabolomes in urine. The mass spectrometer we use works without sample preparation and registers volatile organic compounds in air at room temperature without changing the pH of the sample, i.e. under conditions similar to those in which dogs solve the same problem. The experimental cancer models were male BDF-f1 hybrid mice transplanted with hepatocarcinoma tissue, and similar mice transplanted with healthy liver tissue were used as controls. Our data show that both dogs and our proposed laser mass spectrometry method are able to detect both the entire spectrum of volatile organic compounds associated with the disease and minor changes in this spectrum during its course.

Figure 1

Schematic diagram of the laser mass spectrometer: 1—reflectron time-of-flight mass analyzer (m/∆m = 5000), 2—laser plasma, 3—rotating metal target, 4—diode-pumped Nd:YAG laser (pulse duration 0.5 ns, pulse energy 400 μJ), 5—focusing lens, 6—ionization chamber, 7—microcentrifuge tube.

Figure 2

(a) Time dependence of the total ion current (sample 8379) and (b) cumulative mass spectrum of the sample.

Figure 3

Changes in the tissue mass of hepatocarcinoma H33 (red asterisks) and healthy liver tissue (blue circles) by days after transplantation. The horizontal axis shows days after transplantation, and the vertical axis shows values of the mass of the altered tissue ("tumour") in grams. For clarity, the medians of the same type of mass measurements are connected by dashed lines.

Figure 4

Relative frequency of true positive and false positive responses of dogs to urine samples collected by day. Red and blue markers show the proportions of true positive and false positive reactions of dogs to urine samples, respectively. The shaded areas correspond to the 95% confidence interval for each of the 8 days of study.

Figure 5

Results of applying PCA to mass spectra for three groups by day after tissue transplantation: (A) 1–3 day, (B) 4–5 days, (C) 7–8 days. White circles are urine samples from intact mice, blue circles are urine samples from mice with transplanted healthy liver tissue, red circles are urine samples from mice with transplanted hepatocarcinoma tissue. The urine samples spreading over the days after tissue transplantation are presented in Table 3.