Metabolomics-Based Biosignatures of Prostate Cancer in Patients Following Radiotherapy

Abstract

Metabolomics offers new promise for research on prostate cancer (PCa) and its personalized treatment. Metabolomic profiling of radiation-treated PCa patients is particularly important to reveal their new metabolomic status, and evaluate the radiation effects. In addition, bioinformatics-integrated metabolomics-based approaches for disease profiling and assessment of therapy could help develop precision biomarkers in a context of PCa. We report mass spectrometry-based untargeted (global) serum metabolomics findings from patients with PCa (n = 55) before and after treatment with stereotactic body radiation therapy (SBRT), and intensity-modulated radiation therapy (IMRT) with SBRT, and using parsimony phylogenetic analysis. Importantly, the radiation-treated serum metabolome of patients represented a unique robust cluster on a cladogram that was distinct from the pre-RT metabolome. The altered radiation responsive serum metabolome was defined by predominant aberrations in the metabolic pathways of nitrogen, pyrimidine, purine, porphyrin, alanine, aspartate, glutamate, and glycerophospholipid. Our findings collectively suggest that global metabolomics integrated with parsimony phylogenetics offer a unique and robust systems biology analytical platform for powerful unbiased determination of radiotherapy (RT)-associated biosignatures in patients with PCa. These new observations call for future translational research for evaluation of metabolomic biomarkers in PCa prognosis specifically, and response to radiation treatment broadly. Radiation metabolomics is an emerging specialty of systems sciences and clinical medicine that warrants further research and educational initiatives.

Keywords: biomarkers and diagnostics; mass spectrometry; parsimony phylogenetics; prostate cancer; radiotherapy; untargeted (global) metabolomics.

FIG. 1.

A cladogram based on the parsimony phylogenetic analysis of metabolomics mass spectrometry data. It depicts altered serum metabolome patterns of radiation-treated PCa patients clustering into clades separately from the PCa patient samples prior treatment. The triangles represent clades and subclades of PCa patients before (black) and after radiation (red) treatment. The numbers represent the number of synapomorphies (biomarkers) that delineate each clade, for example, the patients in the fourth clade from the left share 10 unique synapomorphies in addition to the 7 synapomorphies shared among this clade and the fifth clade. Patients in the fifth clade share 71 unique synapomorphies.

FIG. 2.

A metabolic interactome network of altered nitrogen metabolism in response to radiation treatment of localized PCa generated by ConsensusPathDB. Yellow boxes indicate compound and family entities, respectively, light green interaction node color indicates biochemical reaction, and red frames show common interaction molecules in nitrogen metabolism pathway in the aforementioned databases: 2-oxoglutarate (2OG), L-glutamate (L-Glu), L-Glutamine (L-Gln), carbamoyl phosphate (CAP). Dark green edges are from KEGG, dark blue is for Reactome, Spike, dark pink is for Humancyc, light pink is for INOH, yellow is for Netpath, and light green is for Biogrid, Pdzbase, Pdb (http://cpdb.molgen.mpg.de/).

FIG. 3.

A metabolic interactome network of altered porphyrin metabolism in response to radiation treatment of localized PCa patients generated by ConsensusPathDB. Yellow and blue boxes indicate compound and family entities, respectively, light green interaction node color indicates biochemical reaction, and red frames show common interaction molecules in porphyrin metabolism pathway in the aforementioned databases: porphobilinogen (PBG), bilirubin monoglucuronide (BMG), uridine-5-diphosphate (UDP), bilirubin diglucuronide (BMG), bilirubin (BIL), hydroxymetilbiliane (HMBL). Dark green edges are from KEGG, dark blue is for Reactome, Spike, dark pink is for Humancyc, light pink is for INOH, yellow is for Netpath, and light green is for Biogrid, Pdzbase, Pdb (http://cpdb.molgen.mpg.de/).

FIG. 4.

A metabolic interactome network of altered purine metabolism in response to radiation treatment of localized PCa generated by ConsensusPathDB. Yellow and gray boxes indicate compound and family entities, respectively, light green interaction node color indicates biochemical reaction, and red frames show common interaction molecules in purine metabolism pathway in the aforementioned databases: pyruvic acid, xanthosine, inosine monophosphate (IMP), hypoxanthine (Hyp), 2-amino-6-hydroxypurine (Gua), xanthosine-5-monophosphate (XMP), diphosphoric acid, phosphoenolpyruvic acid (PEP), inosinic acid (IMP), guanosine monophosphate (GMP), ATP, ADP. Dark green edges are from KEGG, dark blue is for Reactome, Spike, dark pink is for Humancyc, light pink is for INOH, yellow is for Netpath, and light green is for Biogrid, Pdzbase, Pdb (http://cpdb.molgen.mpg.de/).

FIG. 5.

A metabolic interactome network of altered glycerophospholipid metabolism in response to radiation treatment of localized PCa generated by ConsensusPathDB. Yellow and gray boxes indicate compound and family entities, respectively, light green interaction node color indicates biochemical reaction, and red frames show common interaction molecules in glycerophospholipid metabolism pathway in the aforementioned databases: cardiolipin, inositol (INS), cytidine diphosphate diacylglycerol (CDP-DAG), phosphatidyl serine, 1-acyl-LPL, CDP-glycerol, fatty acid, 2-acyl-LPC, 2-acyl-LPE, phosphatidylglicerophosphate (PGP), 1-acyl-LPG, monolysocardiolipin (MLCL), dihidroxyacetone phosphate (DHAP), glycerophosphoetanolamine (GPETA), glycerophosphocoline (GPCho). Dark green edges are from KEGG, dark blue is for Reactome, Spike, dark pink is for Humancyc, light pink is for INOH, yellow is for Netpath, and light green is for Biogrid, Pdzbase, Pdb (http://cpdb.molgen.mpg.de/).