|
|
|
|
CIVM research
|
|
|
G. Allan Johnson, PhD, GP Cofer, B Fubara, S Gewalt, L Hedlund
(Center for In Vivo Microscopy), B Guenther (MRPath, Inc.), R Maronpot (NIEHS)
|
MR histology of the C57BL mouse
|
Information edited from poster presented at 17th European Experimental Nuclear Magnetic Resonance Conference/32nd AMPERE, Lille, France, September 6-11, 2004, poster 380.
 Introduction
Sequencing the human genome has stimulated interest in mouse models to study human disease—generating a need for methods to effectively phenotype these new models. We use an integrated program for standardized phenotyping of mouse models using MR histology [1]. We acquire 3D arrays of whole mice at 50-micron isotropic resolution. We remove organs/tissue samples and rescan at spatial resolution up to 20 x 20 x 20 microns [2]. The method shown here applies to knockout mice.
|
| Methods
Reference [1] describes the method of active staining for whole mouse perfusion The animal is placed on the heated cradle (Fig 1). Figure 3 shows a schematic of the approach where the animal is perfused with a saline flush, followed by a mixture of 10% buffered formalin and contrast agent (Prohance, Bracco Diagnostics).
Fig 1. An aluminum plate controls temperature during perfusion.
|
Fig 2. Imaging was performed on a horizontal 7.0T magnet with a prototype GE EXCITE console.
|
Fig 3. Catheters were placed in the right jugular and left carotid. 3 steps allow perfusion of the thorax, head, arterial and venous sides of the abdomen.
|
| Results
Fig 4. A representative coronal slice from the isotropic 50-micron array demonstrates three areas of interest (magnified on bottom).
|
Fig 5. MR histology can now apply to phenotyping some transgenic and knockout models. The wild type (left) shows a) liver lobules, b) stomach, c) renal cortex, d) medulla,
e) papilla, f) collecting structures. The matched image from a uricase knockout mouse shows g) substantial alteration of the renal papilla, and h) polycystic lesions in the renal cortex.
|
|
Seminiferous tubules in the testis @ ~100 microns |
|
Villi in the small intestine |
|
Soft tissue resolution is sufficient to differentiate red and white pulp in the spleen and mucosal lining of the stomach |
|
Fig 6. Since MR histology is non-destructive, the same specimen can be scanned multiple times. This kidney was removed and rescanned at higher resolution (25 x 25 x 25 microns). Conventional histology or immunochemistry can also be used on the specimen.
|
Conclusion
MR imaging has fundamentally altered how clinical medicine is done today. With an increase in spatial resolution, MR microscopy promises a similar impact on the basic sciences. Spatial resolution of a routine clinical study (voxels 1 x 1 x 5 mm [5 mm3]) compared to Figure 6 is nearly 300,000X.
The advent of high-temperature superconducting detectors promises sufficient sensitivity to increase resolution to 10 microns in the future. We believe the integration of all of these advances, along with an infrastructure to handle massive amounts of data, will have a profound impact on our ability to interpret the human genome through the use of mouse models.
References
- GA Johnson, et al., Radiology, 222(3): 789-793, 2002
- GA Johnson et al., J Magn Reson Imaging, 16(4):423-9, 2002
|
|
|
|
