Imaging techniques reveal key indicator of metastatic cancer
Actin is one of the primary structural proteins in cells and can exist in a number of configurations and filament types. During cancer metastasis, cells undergo a process called the Epithelial Mesenchymal Transition (EMT), which induces the formation of thick actin stress fibers. These fibers aid the cancer cell in navigating the crowded tumor environment, so their formation could be used as an indicator for metastatic cancer.
Basu et al. describe a collection of analysis methods for studying actin filaments and how they can be used to identify the EMT and metastasis.
“This review primarily focuses on what kind of information can be extracted either directly or indirectly from studying actin,” said author Arkaprabha Basu. “These methods can be a huge source of unmined data, such as the location of early actin expression, distribution of expression throughout the cell, and co-localization with other proteins.”
Several analysis methods, such as western blots, cryo-electron microscopy, and fluorescence spectroscopy, are commonly used to study cell structure and dynamics. These methods are powerful and useful, but the resulting images cannot, on their own, be used to detect the subtle changes in actin stress fibers indicative of metastasis.
With the advent of more advanced image processing and machine learning techniques, the data from these experimental methods can be quantified to highlight changes in actin stress fibers and detect metastasis. The authors discussed several such techniques and the increased potential they can provide.
“With the advent of imaging techniques and artificial intelligence-based image analysis methods, the amount of extractable information has increased,” said Basu. “We intend to give a taste of that potential.”
Source: “The actin cytoskeleton: Morphological changes in pre- and fully developed lung cancer,” by Arkaprabha Basu, Manash K. Paul, and Shimon Weiss, Biophysics Reviews (2022). The article can be accessed at https://doi.org/10.1063/5.0096188 .