by Josh P. Roberts
Scientists have been peering inside live animals at least since Roentgen discovered and tamed the x-ray. These days many and varied modalities are used for in vivo imaging – not only ionizing radiation from x-rays and radiotracers, but also sound, magnetic moment, and even light.
Each has the potential to add its own data set, and comes with its own set of drawbacks as well. To ascertain the shape or location of anatomical structures, for example, a CT scan, MRI, or ultrasound might be the method of choice -- yet these are unlikely to give much information about the where injected tumor cells might have landed, or what surface molecules they now express.
Optical imaging – specifically bioluminescence and fluorescence -- has emerged over the past decade or so as a powerful method for functional in vivo imaging, due to its relatively low cost and ease of use, high sensitivity, the ability to track a cohort longitudinally, minimal toxicity, and a variety of other factors both historical and practical.
Not the least of these is the accessibility of probes – conjugated fluorescent and bioluminescent reporters as well as genetically-engineered tags – making optical imaging “the most widely utilized modality for preclinical molecular imaging,” writes University of California, Davis’ Simon Cherry in a recent review.1
Optical imaging has the potential to establish the efficacy of drug compounds in an animal model, long before any morphological changes might be detected by structural imaging. It also holds promise to establish biomarkers of disease in animals; knowledge that can ultimately be translated to the clinic.
Simplicity itself
The principle behind bioluminescence and fluorescence is surprisingly simple. In the case of the former, it’s generally a light-emitting moiety that is detected – think fireflies and glowing sea creatures. These are most commonly the result of a reaction between an enzyme (a luciferase) and its substrate (a luciferin), creating an electronically-excited state that produces light as it decays.
Cells and even animals can be engineered to produce a luciferase. Other co-factors like ATP and calcium as well may be required for the reaction to proceed, making bioluminescence a powerful mode of detection in a wide variety of assays.
Fluorescence occurs by an analogous process to bioluminescence. Only in this case a molecule (called a fluorochrome, or fluor for short) becomes excited in response to light of a different wavelength – think phosphorescent watches, only of much shorter duration.
In vivo imagers
The effect of promoters driving luciferase-transgenes or fluorescently-tagged antibodies and pharmaceuticals can be imaged in living animals using the same principles – and sometimes in the selfsame machines. Following intravenous injection of luciferin or a pharmaceutical, the reporters can be viewed in one of the several optical imaging instruments now on the market.
After injecting luciferin into the transgenic mice, for example, the animals are placed in the light-tight chamber of a bioluminescence imager, and light is detected where the luciferin-luciferase reaction occurs. Similarly, fluorescence imagers can follow the injected antibodies and drugs as they find their targets or are cleared from the body. Sophisticated detectors and software translate signals emanating from within the animal into an image on a screen.
Some imagers can accommodate up to five animals on a heated bed, with a continuous flow of anesthesia fed through nose cones, allowing the signal to be followed over time.
And because the data is collected vitally, longitudinal studies – not only over the course of minutes, but over hours, days, or even months – can be performed on the same cohort.
What’s the difference?
Yet all optical imaging instruments are not created equal. There are a variety of parameters to consider, notes the director of a major university’s in vivo imaging facility, and there are tradeoffs. Each system offers its own set of ways of producing, acquiring, and processing data. And “they’re vastly different in what they do well and what they don’t do well.”
Bioluminescent imagers, because they don’t require an exogenous light source, and because mammals exhibit little or no endogenous bioluminescence, have extremely little background noise to contend with. They thus have the potential for very high sensitivity. Yet they are limited by the researcher’s ability to engineer or inject an appropriate reporter, and so fluorescence may be the modality of choice.
For fluorescence imaging, though, the fluor needs to be excited by an exogenous light source. This can be from a laser or white light source, and it can come from the same side of the animal as the detector, or the opposite.
By offering the option of using trans-illumination – where the excitation source and detector are on opposite sides of the animal -- Caliper Life Sciences’ IVIS instruments gain an enormous advantage in signal:noise, claims Mark Roskey, the company’s VP for reagents and applied biology.
Yet at the same time, because visible, and to a lesser extent near-IR, light is strongly absorbed by tissue, there are limits to the size of an animal able to be imaged by such an approach, as well as caveats about attempting to image through (or near) hemoglobin-dense organs such as the heart, spleen, and liver.
eXplore Optix system, built by Advanced Research Technology (ART) and marketed worldwide by GE Healthcare, uses a laser pulse every 100 picoseconds to excite the fluors. It then collects not only emission wavelength data, but information on its timing as well. This “lifetime analysis” allows researchers to understand certain processes going on within the animals, says Pierre Couture, ART’s VP of sales and marketing for Optix. The shape of the decay curve of some fluors changes with the environment – pH, whether they’ve crossed the blood-brain barrier, or whether they’re bound or unbound, for example. The system, though, costs several-fold more than other optical imaging instruments.
The CRi Maestro system’s claim to fame is a combination of electronically tunable filters and multi-spectral analysis. Thus it excels in resolving signals from overlapping fluors and from autofluorescence, allowing simultaneous imaging of multiple fluors.
And Kodak’s In-Vivo Imaging System FX Pro offers an integrated CT scanner for multi-modal imaging, allowing for functional (molecular) and structural (anatomical) information from the same machine: “With our system, they can do fluorescence, luminescence, and then by simply sliding a phosphor screen underneath the animal (without touching the animal or moving any of the optical settings), they can take an x-ray image,” explains William McLaughlin, director of R&D for Kodak molecular imaging systems.
For any virtue claimed by one system, another is likely to claim a different way of accomplishing the same task, or downplaying its importance.
IVIS, says Roskey, uses “spectral unmixing” to “deconvolute” the data, accomplishing what amounts to multi-spectral analysis.
Both the IVIS and eXplore Optix systems offer a virtual version of tomography (although Couture isn’t happy with it just yet, noting, “We’re coming out with a new software version that will be available in Q2, that will do 3-D reconstruction.”)
And Couture also points out that the eXplore Optix is designed to work in conjunction with GE Healthcare’s eXplore Locus CT scanner, and the images can be overlaid using readily-available fusion software. (Meanwhile, McLaughlin claims the virtue of Kodak’s not needing any such software.)
The future?
What they don’t do now, they will soon, or so it seems. Kodak, for example, is looking into doing 3D analysis and multi-spectral imaging (the latter of which we should look for this summer, McLaughlin says).
In general, the industry is hoping to solve what might be optical imaging’s biggest issue – depth of penetration – to allow the imaging of larger animals like dogs and monkeys, and eventually humans. This will be accomplished, Roskey says, through a combination of brighter, more specific dyes, as well as some engineering tricks in terms of design of the instruments themselves.
1SR Cherry, Multimodality in vivo imaging systems: twice the power or double the trouble?, Annu Rev Biomed Eng, 8(2006)35-62.
