In Star Trek, Dr. McCoy determined a patient's prognosis with the wave of a portable scanner. However, many current imaging techniques require radioactive or magnetic labels, making them less suitable for quick diagnostic use in humans. Now Manganas et al. report a noninvasive technique to track neural progenitor cells (NPCs) in vivo in a recent article in Science.
Proton nuclear magnetic resonance spectroscopy (1H-NMR) detects small quantities of metabolites in solution. The related technique, proton magnetic resonance spectroscopy (1H-MRS), analyzes these metabolites in live animals. 1H-NMR shows distinct spectra for N-actetyl aspartate (NAA) and choline, which are biomarkers for neurons and glia, respectively.
The authors searched for a 1H-NMR-detectable biomarker for NPCs. Floating aggregates of NPCs from embryonic mouse brain (neurospheres) showed a metabolite at a frequency of 1.28 parts per million (ppm) that was absent in pure populations of neurons and astrocytes. Relative to neurons and astrocytes, neurospheres showed reduced levels of NAA and choline, respectively. Although oligodendrocyte precursor and embryonic stem cells were positive for the 1.28-ppm biomarker, they expressed the biomarker at much lower levels than did neurospheres, suggesting its specific enrichment in NPCs.
Levels of the 1.28-ppm biomarker decreased, and levels of neural and glial markers increased in response to differentiating conditions in cultured neurospheres and in postnatal relative to embryonic mouse brain cells. The adult hippocampus retains NPCs. In the adult mouse, hippocampus contained more 1.28-ppm biomarker relative to cortex, and electroconvulsive shock, which induces neurogenesis, increased the 1.28-ppm biomarker in mouse hippocampus.
What is the 1.28-ppm biomarker? In 1H-NMR, the profiles for saturated and monounsaturated fatty acids overlapped with the 1.28-ppm biomarker profile, and the biomarker declined in neurospheres treated with the fatty acid synthesis inhibitor cerulenin, suggesting that it is a lipid.
1H-MRS should detect the NPC biomarker in the live rat. However, Fourier transform signal processing, which researchers traditionally use to process magnetic resonance imaging (MRI) signals, did not isolate the 1.28-ppm biomarker in the rat brain. Singular value decomposition (SVD), which is a more sensitive signal processing algorithm that helps resolve signal components at low signal-to-noise ratios resolved the 1.28-ppm biomarker. Consistent with 1H-NMR, 1H-MRS showed increased levels of the 1.28-ppm biomarker in hippocampus relative to cortex and in electroconvulsive shock-treated relative to control rats. Both Fourier transform and SVD algorithms resolved the 1.28-ppm biomarker in rats transplanted with NPCs. Relative to vehicle-injected control regions, brain regions containing transplants showed 35 times more 1.28-ppm biomarker.
In people, MRI with SVD-based analysis resolved the 1.28-ppm biomarker in hippocampus but not cortex. Consistent with the age-related decline in neurogenesis, adults 30-35 years of age showed lower levels of the 1.28-ppm biomarker relative to children 14-16 and 8-10 years of age. Perhaps the authors' technique could be used to track disease-related changes in neurogenesis in people.