Fluorescence, Raman scattering, and multispectral imaging enable molecular-level disease diagnosis.
A stream of photons races through the air, aimed at a biological sample. Upon impact, the interaction between light and matter rattles the molecular bonds, and altered light emerges—all in a…well…in a flash of light. Using just the right wavelength of illumination and detecting the results in different ways can reveal the molecular line between health and illness.
Some forms of interrogating molecules with light started long ago. Around 1900, for example, August Köhler—a German scientist who went to work at Carl Zeiss (Jena, Germany)—developed the first ultraviolet microscope, but the autofluorescence of biological materials blinded parts of the view. Nonetheless, this work eventually spawned fluorescence microscopy, in which light of a particular wavelength stimulates the emission of a different wavelength from a specific marker, such as green fluorescent protein. An ever-increasing family of such molecular markers makes fluorescence microscopy one of today’s standard laboratory tools.
|FIGURE 1. Developing diagnostics from surface-enhanced resonance Raman scattering (SERRS) requires several precise processes, including creating the proper substrate, such as the shiny silicon wafer shown here. (Image courtesy of Renishaw Diagnostics)|
Beyond the on or off—it’s there or it’s not—message from fluorescence microscopy, other tools generate more information, delving deeper into the molecular nature of a sample.
Sensing more scattering
In Raman scattering, single-wavelength laser light causes a sample’s bonds to vibrate and stretch. The resulting scattered light comes off as a collection of wavelength peaks, essentially providing a molecular fingerprint of the sample. “In Raman scattering,” says David Eustace, Ph.D., project leader for development at Renishaw Diagnostics (Glasgow, Scotland), “every peak can be related to a feature in the molecule.”
To magnify this effect, researchers bind the molecule to a roughened metal surface. “This acts like an antenna that focuses the light more effectively into the target,” says Eustace. The laser light excites surface plasmons—oscillations of electron density—in the metal substrate, and the plasmons interact with the absorbed molecule (see Fig. 1). In essence, the molecule experiences a localized electric-field enhancement, which amplifies the intensity of the Raman effect. This so-called surface-enhanced Raman scattering can magnify the signal by a million times, but it’s still not enough for molecular biology.
To go even further, Renishaw Diagnostics adds a chromophore—a dye—to the sample. In such surface-enhanced resonance Raman scattering (SERRS), matching the chromophore’s absorption of light with the wavelength of the laser light magnifies the signal by a trillion times or more. As Eustace describes this: “It’s a massively amplified Raman signal, which delivers a real increase in sensitivity.”
|FIGURE 2. The RenDx SA-1000 analyzer uses surface-enhanced resonance Raman scattering (SERRS) to allow multiplexing of reactions in search of as many as 10 markers at once. (Image courtesy of Renishaw Diagnostics)|
Multiplexing the markers
By using multiple dyes attached to different molecules, this technique can simultaneously identify them (see Fig. 2). For example, the dyes can be added to DNA, which can be isolated with magnetic beads. The resulting labeled chunks of nucleic acids can then be analyzed with SERRS.
“We can do 10 dyes in one well of a plate,” says Alastair Ricketts, Ph.D., principal scientist at Renishaw Diagnostics. “So in a disease state, we could examine multiple causes in one assay instead of doing many tests.” So Ricketts and his colleagues hope to apply this technology as a diagnostic for diseases that can arise from one or more causative agents. For example, it could reveal a multiple-virus infection or a dual viral and bacterial infection. “We’re focused on infectious diseases for now,” says Ricketts.
To run such an assay based on nucleic acids, DNA or RNA, researchers spend about an hour extracting the nucleic acids from the sample. Then, the nucleic acids must be amplified using the polymerase chain reaction (PCR), which takes a couple of hours. Separating the DNA of interest takes a couple more hours. Then, a complete 96-well plate can be read in just 15 minutes or so. “It’s a turnaround time of a single day, which was our goal,” says Ricketts.
Beyond nucleic acid-based tests, scientists at Renishaw Diagnostics also envision other approaches. For example, they are working on antibody-based assays that would look for specific proteins in a sample.
For diagnostic assays, this company’s team is running evaluations, letting independent labs validate the technology. After that, the company hopes to move toward clinical testing in Europe and the U.S. A research-use-only platform is already available.
|FIGURE 3. AQUAnalysis software can combine a range of molecular structures, such as the cytokeratin proteins (green) in the cytoplasm and DNA (blue) in the nucleus labeled here. (Image courtesy of HistoRx)|
Given today’s breadth of technology that could enhance molecular diagnostics, teamwork makes good sense in some cases. At Caliper Life Sciences (Hopkinton, MA), for instance, Darren Lee, vice president – marketing, cellular and tissue analysis strategic business unit, says, “Our whole focus is on imaging proteins, including quantifying expression from in-tissue samples.” To do that, they develop imaging systems that analyze a sample with light across a spectrum of wavelengths. In August, Caliper Life Sciences launched a version of its Vectra—an automated, multiplexing, tissue-imaging platform—that includes the AQUAsition and AQUAnalysis software from HistoRx (Branford, CT; see Fig. 3). “This will provide even more accuracy in analyzing images,” Lee says.
In describing the Vectra’s capabilities, Cliff Hoyt, divisional vice president of tissue applications and collaborations at Caliper Life Sciences, points out that a liquid-crystal tunable filter in the microscope’s optical path allows the device to image a sample at multiple wavelengths. “With the resulting multispectral image cube,” says Hoyt, “you can reliably and accurately isolate the fluorescent label’s emissions from autofluorescence.”
The Vectra first uses its spectral unmixing capabilities, which are followed by AQUA’s analysis. “This delivers a hands-free, automated assessment of the protein expression,” says Hoyt. In addition, the AQUA-enhanced Vectra brings advanced pattern recognition to molecular diagnostics. “Part of the AQUA system uses markers to mask areas of interest,” Hoyt explains.
As one example, this system can be used in digital pathology. Consequently, a pathologist could quantify disease-related proteins in a sample to diagnose an illness.
“In today’s immunohistochemistry,” Lee says, “pathologists use visual assessments, but it’s difficult to judge intensity.” He adds, “HistoRx has shown that automated, quantitative analysis of the proteins provides more precision.” In fact, this analysis might even diagnose who would respond to, say, a cancer drug and who would not.
For the moment, the HistoRx software only tracks one target at a time. “The next step is to look at 2, 3, 5 markers in a clinical setting,” says Hoyt. “Then, you can ask even more demanding questions.”
So from scattered photons to multispectral approaches, new technologies keep unveiling more about molecules that can diagnose diseases. These techniques could even lead physicians to treatments targeted for specific individuals.