super(1)H-NMR metabonomics analysis of sera differentiates between mammary tumor-bearing mice and healthy controls

Global analysis of super(1)H-NMR spectra of serum is an appealing approach for the rapid detection of cancer. To evaluate the usefulness of this method in distinguishing between mammary tumor-bearing mice and healthy controls, we conducted super(1)H-NMR metabonomic analyses on serum samples obtained from the following: 10 mice inoculated with a highly-metastatic mammary carcinoma cell line, 10 mice inoculated with a "normally" metastatic mammary carcinoma cell line, and 10 healthy controls. Following standard spectral processing and subsequent data reduction, we applied unsupervised Principal Component Analysis (PCA) to determine if unique metabolic fingerprints for different categories of metastatic breast cancer in serum exist. The PCA method correctly separated sera of tumor-bearing mice from that of normal healthy controls, as shown using the scores plot which indicated that sera classes from tumor-bearing mice did not share multivariate space with that from healthy controls. In addition, this technique was capable of distinguishing between classes of varying metastatic ability in this system. Metabolites apparently responsible for separation between diseased and healthy mice include lactate, taurine, choline, and sugar moieties. Results of this study suggest that super(1)H-NMR spectra of mouse serum analyzed using PCA statistical methods indicate separation of tumor-bearing mice from healthy normal controls, justifying further study of the use of super(1)H-NMR metabonomics for cancer detection using serum.