Microfluidics, fluorescence could yield 3D tissue maps for improved disease diagnosis

As biopsies are limited in their ability to measure the protein levels necessary for accurate disease diagnosis, a team of researchers at the University of Notre Dame (Indiana) has developed a method involving microfluidics and fluorescence to acquire 3D maps of tissue that provide much more information, incorporating both data on the tissue structure and its molecular profile.

In conventional methods, pathologists typically take thin, two-dimensional slices of the tissue and use hematoxylin and eosin (H&E) to stain the cells. Hematoxylin turns nuclei blue while eosin turns other cellular components pink, enabling doctors to see cellular structure and identify signs of cancer—the gold standard for diagnosing cancer and other diseases, according to Jeremiah Zartman, an assistant professor in the Department of Chemical and Biomolecular Engineering at the University of Notre Dame who was involved in the work. However, a flat image cannot reveal anything about the shape and curvature of the tissue, and the process unavoidably destroys some of the sample. Sometimes there is not much of the sample, and doctors may want to limit the amount because biopsies can be painful for the patient.

Two-dimensional slices also limit how much information can be extracted. If clinicians or researchers want to identify the location of specific proteins by tagging them with a dye, for example, they can only do so for a few at a time because of the limitations of fluorescence microscopes. If they want data on additional proteins, they would need yet another slice. But the research team's microfluidic approach involves placing a sample inside a tiny chamber on a clear chip not much bigger than a dime. The enclosure lets the tissue maintain its shape and structure, which enables multiple rounds of staining and imaging. Narrow channels allow chemical solutions to be injected into the tissue.

This image shows a stained kidney sample imaged on a two-photon microscope with blue representing nuclei and red cytoplasm. (Image credit: Cody Narciso and Victoria Zellmer/University of Notre Dame)

Those solutions enable doctors or researchers to peer into the tissue. First, they have to remove lipids, a type of fatty molecule that makes the tissue opaque. Then, they can use fluorescent dyes to tag specific proteins that could be markers for diseases. Because the sample is preserved in the chip, researchers can clear those stains, using a method called quenching, and tag another set of proteins. In principle, Zartman says, researchers can repeat this process and produce a rainbow of colors that identify the relative amounts of dozens of proteins within cells throughout the whole tissue.

The chip also provides long-term storage of the sample, allowing for analysis at a much later time. Its compact size requires smaller amounts of chemicals, which could potentially cut costs. Future designs will include control devices to automate the staining analysis. The doctor or researcher would simply take the chip and insert it into their instruments, turn it on, and collect the data.

So far, the research team has done a proof-of-principle demonstration. Their experiments, in which they used different stains on multiple kinds of mice tissue, show that the 3D images can be made into slices that qualitatively match conventional images of biopsies—making the transition to the new technique easier. The researchers are now modifying the process to work with other archived tissues, developing an automated process, and planning to stain more proteins at once to provide a detailed map that conveys both cell shapes and protein levels within the tissue.

Full details of the work appear in the journal Biomicrofluidics; for more information, please visit http://dx.doi.org/10.1063/1.4941708.