Profiling cancer metabolism at the ‘omic’ level: a last resort or the next frontier?
© Gelman and Patti. 2016
Received: 5 February 2016
Accepted: 5 February 2016
Published: 21 March 2016
When profiling metabolites, there are two general experimental paradigms: untargeted studies at the omic scale and targeted studies which usually focus on a much smaller subset of compounds. Although untargeted studies have become increasingly fashionable and are often perceived to be the cutting edge, it is important to recognize that they have limitations and are not always the better experiment. In theory, an untargeted analysis profiles all of the same metabolites as a targeted analysis, plus more. So why is more not always better?
We will frame our consideration in the words of pioneering computer scientist Alan Perlis: “Fools ignore complexity. Pragmatists suffer it. Some can avoid it. Geniuses remove it.” .
“Fools ignore complexity.”
Indeed, untargeted metabolomic datasets are exceedingly complex. When applying a mass spectrometry-based platform, biological samples typically generate tens of thousands of signals per experiment. Even with state-of-the-art technologies, the majority of these signals cannot be annotated. Some signals are challenging to annotate because they are experimental artifacts, while others correspond to metabolites whose structure, function, and pathway remain unknown . Without annotation, it is challenging to ascribe global meaning to the datasets. For example, it is precarious to compare the global metabolism of two samples on the basis of the percentage of signals changing because this percentage is highly dependent upon the number of artifacts.
“Pragmatists suffer it.”
Despite its challenges, untargeted metabolomics is the workflow of choice for some research applications, such as identifying biomarkers of disease. In these applications, researchers do not have a biochemical hypothesis prior to beginning the analysis and generally only attempt to identify metabolomic signals that are potentially diagnostic of the condition being studied. A convincing example of the power of untargeted metabolomics to identify serum markers of non-small-cell lung cancer was recently published in the Journal of Clinical Oncology. Wikoff et al. found that the concentration of diacetylspermine increases approximately twofold in the blood collected from patients 6 months before diagnosis . Untargeted metabolomics is also well suited for other types of unbiased screening applications. In Cancer and Metabolism, for instance, Gelman et al. reported an application of untargeted metabolomics to find metabolic products of the oncometabolite 2-hydroxyglutarate .
“Some can avoid it.”
“Geniuses remove it.”
Great progress has been made over the last decade to improve the robustness of the mass spectrometry-based metabolomic platform and to facilitate the associated data processing. Instruments are becoming more quantitatively reliable, databases are expanding, software is advancing, computational approaches to assess flux are evolving, and informatic strategies to integrate gene expression data are being developed. Some of these advances are reviewed in Cancer and Metabolism by Markert et al. in “Mathematical Models of Cancer Metabolism” and in “Integration of Omics: More than the Sum of its Parts” by Buescher et al. [6, 7]. The objective is not only to reduce complexity but also to increase accessibility so that integrated omic experiments can be performed by scientists without extensive technical expertise.
This leaves us with an interesting question: should untargeted metabolomics be reserved as a last resort when no other experimental approaches can solve the problem of interest? We suggest that the answer is yes. However, the problems that metabolomic technologies uniquely position us to solve are of exceptional importance to the field of cancer metabolism and, in this sense, represent the next frontier.
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.
- Perlis AJ. Epigrams on programming. ACM SIGPLAN Notices. 1982;17(9):6.View ArticleGoogle Scholar
- Mahieu NG, Huang X, Chen YJ, Patti GJ. Credentialed features: a platform to benchmark and optimize untargeted metabolomic methods. Anal Chem. 2014;86 (19):9583–9589
- Wikoff WR, Hanash S, DeFelice B, Miyamoto S, Barnett M, Zhao Y, et al. Diacetylspermine is a Novel prediagnostic serum biomarker for non-small-cell lung cancer and has additive performance with pro-surfactant protein B. J Clin Oncol. 2015;33(33):3880–6.View ArticlePubMedGoogle Scholar
- Gelman SJ, Mahieu NG, Cho K, Llufrio EM, Wencewicz TA, Patti GJ. Evidence that 2-hydroxyglutarate is not readily metabolized in colorectal carcinoma cells. Cancer Metab. 2015;3:13.View ArticlePubMedPubMed CentralGoogle Scholar
- Salamanca-Cardona L, Keshari KR. (13)C-labeled biochemical probes for the study of cancer metabolism with dynamic nuclear polarization-enhanced magnetic resonance imaging. Cancer Metab. 2015;3:9.View ArticlePubMedPubMed CentralGoogle Scholar
- Markert EK, Vazquez A. Mathematical models of cancer metabolism. Cancer Metab. 2015;3:14.View ArticlePubMedPubMed CentralGoogle Scholar
- Buescher JM, Driggers EM. Integration of omics: more than the sum of its parts. Cancer Metab. 2016.