Abstract

Introduction

Materials & Methods

Results

Discussion & Conclusion

References

Discussion
Board

Animals - Table 1 gives the age and the species of the specimens examined. The mouse lemurs (Microcebus murinus) were obtained from the Duke University Primate Center. All animals died of natural causes. The housing conditions and the diet of the dwarf lemurs (Cheirogaleus medius) were similar to the ones described in Lee et al. (1996). The mouse lemurs were fed cracked Purina Old World Monkey Chow (nr. 5038), chopped fruits, vegetables and crickets. The chow contains 180 parts iron per million. There was no supplementation with vitamins or minerals.

Table1

Species and age of specimens examined

Species                Age (yr)           Gender
dwarf lemur             15                   male

mouse lemur             12                   female

mouse lemur              8                   male

mouse lemur              8                   female

dwarf lemur              3                   male

mouse lemur              2                   male

mouse lemur              2                   female

Magnetic resonance microscopy - MR microscopic images (Narasimhan et al., 1994; Narasimhan and Jacobs, 1996) of the intact thawed cadaver head were made at 2°C using a Bruker-AMX 500 MHz spectrometer consisting of an X-32 based acquisition system and an 89 mm vertical bore 11.7-Tesla magnet (Bruker Instruments, Billerica, MA). The microscope is equipped with a 16-bit digitizer (sweep width = 125 kHz; sampling time = 8 ms) and a gradient interface board that allows on the fly calculations of gradient arrays. Images were acquired using T2*-weighted 3-D Gradient Echo scheme (Haase et al., 1986; Chien and Edelman, 1991) with TR=100ms, TE=9ms, 30 micron isotropic voxel resolution, and ~4hrs imaging time. A 2-cm-diameter Alderman-Grant rf coil and a shielded gradient coil (30 G/cm maximum gradient) were used for imaging.

Images were acquired under similar conditions on all animals as described by Dhenain et al. (1997). Imaging probe adjus tments were very reproducible, leading to same values of radiofrequency pulse amplitudes in each cases. The only parameter allowed to vary deliberately was the receiver gain, determined automatically at the beginning of image acquisition. After reconstruction on the spectrometer computer, images were tranferred to a SiliconGraphics workstation and analyzed using VoxelView software (Vital Images, Fairfield, IO). After correction for different receiver gains, images from different animals could be directly compared (Dhenain et al., 1997).

Data analysis - A description of imaging protocol has been previously reported (Dhenain et al., 1997). Signal intensity was measured using a gray scale ranging from 0 (black) to 255 (white). We localized the brain structures with atlases of the mouse lemur brain (Bons et al., 1998; Cooper, unpublished). We then assessed the mean-intensity values of globus pallidus, substantia nigra and basal forebrain structures. The contours of the structures were delineated o n the right and left hemispheres and the two corresponding mean-intensity values were averaged.

A high correlation between the mean-intensity values measured by two different observers has been observed by Dhenain et al. (1997) for several brain structures including the globus pallidus and the substantia nigra. A maximum threshold was applied in order to set a range of voxel values to be displayed. This threshold corresponds to the mean-intensity values of globus pallidus, substantia nigra and basal forebrain structures for each specimen. In the case of the specimen illustrated here, the upper threshold was set at 35 (figs. 1-4) and at 41 (fig. 5). The lower voxel threshold was set at 0. Voxel values beyond the upper thresholds are eliminated in the right images of figs. 1-5. To permit the remaining voxels (0-35 and 0-41) to be visualized, the intensity histogram was adjusted to enhance contrast. The resulting images diplay the globus pallidus, substantia nigra, basal forebrain structures as well a s some adjacent structures with similar mean-intensity values.

Histochemistry - Within 24-72 hrs. after MR imaging, the brains were removed and refrigerated in 10% formalin solution with 30% sucrose overnight. Serial sections were cut in sagittal and coronal planes at 50 microns on a cryomicrotome. The sections were kept as free-floating sections (Hill and Switzer, 1984) and divided into three sets of alternating sections. The first set of sections was immersed at room temperature for 30 minutes in Perl's (Prussian) solution (freshly prepared solution of equal parts 2% HCL and 2% potassium ferrocyanide).

The Perl's Blue reaction is a histochemical stain for non-heme iron (Pearse, 1961). This reaction depends on the production of ferric ferrocyanide when ferric ions (Fe 3+) in the tissues react with ferrocyanide in acid solution. It generally reflects the quantity of Fe 3+ stored in the brain parenchyma. We counterstained with pararosaniline for cell nuclei. The second set of sections was pr ocessed with Congo red stain and the third set treated with the Bielschowsky silver impregnation for amyloid deposits and degenerated neurites. The sections were then mounted on gelatin-coated slides. We observed no evidence of amyloid plaques or degenerated neurites in our specimens but we did note the accumulation of lipofuscin granules with epi-fluorescence microscopy in the old animals.

Back to the top.