Figure 1.
1HMRS spectra from the rat hippocampus and cortex analyzed with LCModel.
Fig. 1A: The position of the hippocampal voxel is shown on T2-weighted MRI in the top right corner. The corresponding in vivo 1HMRS spectrum is shown analyzed by LC Model. Fig. 1B: The position of the cortical voxel is shown on the T2-weighted MRI as well as the in vivo corresponding 1HMRS spectrum. The two spectra shown are of high quality with spectral resolution necessary to resolve the labeled metabolites. Specifically, the raw unsmoothed spectra are shown (black),the LCModel fitted spectral output (red solid lines) and the corresponding macromolecular (MM) scan (dotted red lines) measured in the same animal using a PRESS metabolite nulled inversion-recovery pulse sequence in addition to the residual signals (top spectra). tCr = total creatine; Glx = glutamate + glutamine; Ins = myo-Inositol; Tau = Taurine; tCho = total choline; NAA = N-acetyl-aspartate; Glu = glutamate; MM = macromolecules; Lac = Lactate; PG = Propylene glycol.
Figure 2.
Summed and averaged 1HMRS hippocampal spectra before and after ECS.
Fig. 2A: The summed and averaged baseline spectra acquired in the hippocampus from all the ECS rats (n = 7) are shown in red and the summed post-ECS spectra from the same rats are shown in blue. Note that the NAA peak is smaller in the summed post-ECS spectrum when compared to baseline summed and averaged spectra. Fig. 2B: Enlargement of the spectral profiles in the spectral range of 1.5 ppm–1.0 ppm from the summed baseline and post-ECS spectra are shown demonstrating that among the visible peaks, only propylene glycol appears to be different between the two conditions. NAA = N-acetyl-aspartate.
Table 1.
Hippocampal metabolite concentrations before and after ECS or sham treatment.
Figure 3.
ECS does not induce neuronal loss.
Fig. 3A: Representative images of the DG of sham and ECS-treated rats immunostained with neural marker NeuN. Fig. 3B: Representative images from sham and ECS-treated rats stained with anti-NeuN, anti-activated caspase 3, propidium iodide, and Hoechst33342. Arrows show activated caspase 3-positive apoptotic cells. Apoptotic cells stained for activated caspase 3-positive are characterized by compacted and shrunken nucleus as accessed by Hoechst33342 and PI. Fig. 3C: Histogram illustrating that there is no difference in the number of activated caspase 3-positive cells in the DG from sham (n = 8) and ECS (n = 7) rats. Fig. 3D: Distribution of activated caspase 3-labeled cells in the DG, illustrating that the hilus and inner molecular layer contains the majority of apoptotic cells. The SZG had very few apoptotic cells and the granular cell layer (GCL) did not show cells positive or activated caspase 3. Scale bars: Fig. 3A, 100 µm; Fig. 3B, 10 µm.
Figure 4.
Glioblastoma size varies in the RCAS-TVA-J12p16/M9Pten mouse model.
T2-weighted horizontal MRIs at the level of the lateral ventricles from four RCAS-TVA-J12p16/M9Pten mice with large glioblastomas (GBMs) (Fig. 4, A–D) and two RCAS-TVA-J12p16/M9Pten mice who developed smaller frontal GBMs) (Fig. 4, E–F) are shown. For each MRI image, a corresponding color-coded icon is shown to the right illustrating the GBM (black) in relation to ventricles (white) and the rest of the brain (red). As can be seen, the larger GBMs are compressing the lateral ventricles and incorporate both hemispheres. The T2-weighted MRIs of the mice with large GBMs demonstrate that the tumor tissue contain scattered areas associated with lower signal intensity and further that tissue at the edge of the tumor often is associated with brighter signal intensity. The smaller GBMs in Fig. 4E and F are located in frontal cortical areas and also contain tissue with mixed low and bright signal intensities. Scale bar = 2 mm.
Figure 5.
1HMRS spectra from glioblastoma are characterized by increased lipid signals.
The 1HMRS voxel location on a T2-weighted MRI and the corresponding 1HMRS spectra from a control wild type mouse (Fig. 5 A, B) and a RCAS-TVA-J12p16/M9Pten mouse with large a glioblastoma (GBM) (Fig. 5 C, D) are shown. The 1HMRS spectrum from the GBM is dominated by the large and broad lipid peaks at 1.3 ppm and 0.9 ppm. The raw unsmoothed spectra are shown (black), the LCModel fitted spectral output (red solid lines). tCr = total creatine; Glx = glutamate + glutamine; Ins = myo-Inositol; Tau = Taurine; tCho = total choline; NAA = N-acetyl-aspartate; Glu = glutamate; MM = macromolecules; Lac = Lactate; GABA = gamma-aminobutyric acid.
Figure 6.
Immunohistochemistry of glioblastoma with elevated mobile lipid signal.
To assess tumor growth, the RCAS-TVA-J12p16/M9Pten mice were injected with BrdU (150 mg/kg, i.p.) 2 hr before euthanasia. Fig. 6A: T2-weighted MRI of a RCAS-TVA-J12p16/M9Pten mouse who developed a small glioblastoma GBM in the left hemisphere. The region of the cerebral cortex (white inset) containing GBM and normal tissue was tracked and matched with histology (Fig. 6B–E). Fig. 6B: Immunostaining for BrdU and activated caspase 3, demonstrating that the GBM is filled with dividing cells mixed with islets of apoptotic cells. Fig. 6C–E: Higher magnification views of the inset in Fig. 6B illustrating the high density of dividing BrdU-positive cells intermingled with scattered apoptotic cells. (Fig. 6F–H) The GBM tissue immunostained for anti-BrdU and GFAP, a marker of neural stem cells and mature astrocytes. The images presented in Fig. 6B–H are single projection views based on 20-mm serial optical z-stacks. Scale bars: Fig. 6B, 1 mm; Fig. 6C–E, 10 µm; Fig. 6F–H, 100 µm.
Figure 7.
BrdU+ dividing cells and Nile red+ pyknotic cells in glioblastomas.
Fig. 7A: Low magnification images of an RCAS-TVA-J12p16/M9Pten mouse with a glioblastoma (GBM), characterized by high cell density as demonstrated by counter-staining nuclei using PI. BrdU+ dividing cells are clearly evident within the tumor. Fig. 7B: Higher magnification images of the tumor, demonstrating high proliferative activity; BrdU+ dividing cells are distributed throughout the tumor (upper panel). Pyknotic cells are present in the very center of the tumor (lower panel). Pyknotic cells are characterized by PI staining as having smaller and more compact nuclei (lower panel). Scale bars, 10 µm. Fig. 7C: Higher magnification images of the tumor showing Nile red+ lipid droplets (visualized as small densely stained green structures). Nile red+ cells with intracellular lipid droplets (white arrows) and extracellular Nile red+ lipid droplets (red arrowheads) are clearly visible. The perimeter of the tumor (upper panel) is characterized by a smaller number of Nile red+ lipid droplets as compared to the center of the tumor (lower pane) suggesting that the latter contains more necrotic cells. Hoechst33342 (H), a nuclear counter-staining dye. Scale bars, 10 mµ; please note that in Fig. 7B–C representative images of cell density and populations are acquired at 0.5 µm of z stack depth. Fig. 7D: Cell density was determined by Hoechst33342 (H) staining. The tumor-enriched area is characterized by significantly higher cell density when compared to that of control mice or tumor-free tissue. Fig. 7E: Dividing cells were determined by BrdU immunohistochemisty and non-dividing cells by the nuclear counter-staining with PI. About 22% of cells were BrdU+ dividing cells in the tumor-enriched areas; and when correcting for cell density there were 40-fold more BrdU+ cells in the tumor when compared to non-tumor tissue. Fig. 7F: Pyknotic cells with small and compact nuclei are observed in the very center of the tumor. Approximately 5% of cells in the tumor-enriched area were pyknotic (20-fold more dead cells were observed in the tumor when compared to normal tissue). Fig. 7G: When accounting for cell density there were at least10-fold more lipid-containing cells in the tumor when comparing to non-tumor tissue.
Table 2.
Metabolite profile of WT and RCAS-TVA-J12p16/M9Pten with large glioblastomas.