Metabolic Phenotype of the Healthy Rodent Model Using In-Vial Extraction of Dried Serum, Urine, and Cerebrospinal Fluid Spots

High-throughput multiplatform metabolomics experiments are becoming an integral part of clinical and systems biology research. Such methods call for the adoption of robust sample storage and transport formats for small volumes of biofluids. One such format is the dried biofluid spot, which combines small volume requirements with easy portability. Here, we describe ultra high-performance liquid chromatography-mass spectrometry (UHPLC-MS) metabolomics of dried rodent serum, urine, and cerebrospinal fluid spots. An in-vial extraction and UHPLC-MS analysis method was first developed and validated by fingerprinting two test fluids, rat serum and RPMI cell nutrient medium. Data for these extracts were compared in terms of (i) peak area measurements of selected features to assess reproducibility and (ii) total fingerprint variation after data pretreatment. Results showed that percentage peak area variation was found to range between 1.4 and 9.4% relative standard deviation (RSD) for a representative set of molecular features. Upon application of the method to spots bearing serum, urine or cerebrospinal fluid (CSF) from healthy rats and mice, a total of 1 182 and 2 309 reproducible molecular features were obtained in positive and negative ionization modes, respectively, of which 610 (positive) and 991 (negative) were found in both rats and mice. Feature matching was used to detect similarities and differences between biofluids, with the biggest overlap found between fingerprints obtained in urine and CSF. Our results thus demonstrate the potential of such direct fingerprinting of dried biofluid spots as a viable alternative to the use of small (10-15 mu L) volumes of neat biofluids in animal studies.