Digital "Visible Mice" Will Speed Study of Genetic Disease
DURHAM, N.C. -- Advanced imaging technologies have enabled researcher to produce 3-D "magnetic resonance microscopy" (MRM) images of mice at more than 250,000 times greater resolution than MRI scans used to diagnose human disease.
This "Visible Mouse" project will offer a powerful new tool for exploring the morphologic effects of genetically altering mice, believe the Duke University Medical Center researchers. Such mice have become the principal animal model for exploring a vast range of human disorders from cancer to drug addiction. Thus, say the researchers the new MRM technology will open a new era in using mouse models to understand human disease.
They foresee creating MRM images of a multitude of such visible mice, both normal strains and gene-altered animals. Such scans could be transmitted over high-speed Internet 2 connections, enabling researchers to "share" animals electronically, and to digitally "slice" the animals in a variety of ways to explore their anatomy. Thus, a team of researchers separated by thousands of miles could electronically dissect the same animal simultaneously, sharing their insights via videoconferencing.
The researchers, led by G. Allan Johnson, Ph.D., director of the Duke Center for In Vivo Microscopy, described the new technology and its potential scientific impact in an article in the March 1, 2002, issue of the journal Radiology. A technical description of the MRM technology will be published in the June 2002 Journal of Magnetic Resonance Imaging.
In the article, the researchers described a demonstration scan of a mouse in which the gene for the enzyme uricase – an enzyme that converts uric acid in the body -- had been knocked out. In that scan, they achieved a resolution of 110 microns of the whole mice, and could "zoom in" to scan individual organs at a resolution of 25 microns. By comparison, the width of a human hair is about 200 microns.
Basically, MRM, like MRI, involves subjecting animals to high magnetic fields that, while harmless, cause the alignment of protons in the molecules of their tissues. Radio frequency pulses are used to perturb and probe the tissues. By using sensitive radio frequency detectors to map the distinctive response of different tissues to such perturbation, the researchers can create high-resolution images that reveal whole organs as well as the structures within the organs.
The term "Visible Mouse" alludes to the National Library of Medicine's Visible Human project, in which a representative male and female were physically sliced and scanned to produce an anatomically detailed representation of the normal human body.
"With MRM technology, we are at the right place at the right time to help advance the explosion of research on mouse models of human disease," said Johnson. "Physically slicing and staining mouse sections is enormously expensive and severely limits exploration of the often diverse morphological impacts of genetic alteration of these mice. Thus, we believe that we will provide a powerful new tool for molecular biologists, geneticists, toxicologists, pathologists – any scientist using mouse models."
Some 6 million such transgenic and knockout mice were produced for research last year, said Johnson. Transgenic mice are those in which a gene has been altered, and knockout mice are those in which the gene function has been eliminated.
While the Duke center will continue to advance MRM technology, Johnson and colleague Robert Lontz have founded a company, MRPath Inc., to offer MRM scanning services and viewing software to a much larger research community.
According to Johnson, conventional histologic sectioning and staining of mice typically costs about $7 per slice.
"With MRM, we are typically producing as many as 2,600 slices per animal, so it's obvious that conventional slicing and staining a whole animal is prohibitively expensive." What's more, said Johnson, physical slicing unavoidably distorts tissue.
"If a scientist is trying to understand the complicated pathology of a disease, it's necessary to page back and forth among slices," he said. "In the case of histologic sections, no matter how carefully done, there is distortion that prevents one slice from being precisely aligned with another."
Since digital MRM technology is non-destructive, researchers can also use the scans to survey an entire animal, then use traditional methods to explore a particular area of interest in the higher resolution available through conventional optical microscopy, said Johnson. The MRM technology also solves a major problem facing studies of mouse models, in that unexpected pathologies may appear anywhere in an animal.
"With MRM, the researcher can review the whole animal at once, sometimes discovering surprises," he said. "In some of our initial studies of mice, we found that we might be focusing on the kidney, but suddenly discover differences in the thyroid or the heart or the testes. In the case of our scan of the uricase knockout mouse, we detected tumors in places that had not been expected."
Also important, said Johnson, is that the staining chemical perfused through the mouse to increase visibility to MRM leaves water in the tissues.
"Conventional histologic staining leaves the tissue dehydrated," he said. "But as we have seen in clinical use of MRI, how water is bound to tissue has enormous diagnostic importance. Such structures as tumors, inflamed or ischemic tissue and fibrosis all alter how water is bound in the tissue. So, MRM presents the animal in a much more realistic physiological state."
The inherent three-dimensionality of MRM enables researchers to electronically slice the animal in any plane, and even to rapidly step through planes to reveal structures dimensionally, said Johnson.
"For example, a conventional section of a structure such as a papilla wouldn't allow you to accurately measure its size, because you wouldn't know where you had cut the structure," he said. "But with MRM, you could page back and forth until you found the representative area."
Researchers can even create 3-D views of organs or structures, to "fly through" them to explore their morphology, said Johnson.
Finally, the digital nature of the MRM scans also means that researchers can easily download and view comparable slices or 3-D views from a multitude of normal and gene-altered animals. Such comparison enables them to locate the same structural landmarks in different animals.
Johnson and his colleagues will collaborate with the Jackson Laboratory, the largest source of laboratory mice, to create MRM atlases of the major strains if normal mice – both male and female and at a range of ages. These scans will be integrated with the laboratory's database of phenotype information on their mouse strains.
Scans of these normal "visible mice" will be posted on the Web so that researchers can compare their gene-altered strains. Ultimately, Johnson believes, MRM atlases will include a broad range of normal and gene-altered strains, so that researchers can compare the altered mice for common morphological changes.
"For example, if a researcher produces a mouse with altered renal structure, he or she can search the database for other animals with such alterations, and discover whether there are correlations with the animals' genotypes," said Johnson.
Future efforts will also include forging collaborations between MRM researchers and pathologists to interpret the morphological information revealed by MRM, said Johnson.
"We are now in a very similar situation that we were when the first MRI machines began to roll into clinics," he said. "Just as the radiologists and clinicians had to learn the differences between what MRI scans versus CT X-ray scans revealed, now we'll have to understand what is revealed by MRM versus traditional histologic sections. We expect these collaborations will yield a new generation of magnetic resonance histology pathologists."
The new MRM scanning capability has built on technological advances in MRM made over the last 15 years, and supported by the NIH National Center for Research Resources, said Johnson. These advances include more powerful magnets, larger imaging arrays, highly sensitive radio frequency receivers, advanced computer image processing techniques and improved perfusion methods and stains. Practical sharing of the massive amounts of data generated by MRM scans will also depend on the high speed Internet 2, he said.