Colocalization and cancer cells

A rapidly growing evidence suggests the possibility of the link between changes of the degree of colocalization of various antigens in the cells and the spectrum of their malignant properties. This is hardly surprising, since it has been known for a long time that various proteins are overexpressed in cancers. Discordant overexpression of colocalized proteins should inevitably result in the shift of their colocalization properties.

A recent study that caught my attention described overexpression of adenylate cyclase-associated protein 1 (CAP1) in pancreatic cancers. Fluorescence microscopy showed colocalization of CAP1 with actin to the edge of lamellipodia. Knockdown of CAP1 by the RNA interference technique resulted in the reduction of lamellipodia formation, hence colocalization as well, and motility and invasion properties of pancreatic cancer cells. Interestingly, the authors concluded that this is the first report that demonstrated the overexpression of CAP1 in pancreatic cancer cells and suggested the involvement of CAP1 in their aggressive behavior. Thus, the described observation of colocalization of CAP1 with actin served to identify CAP1 as a marker for this type of cancer as well as determined its usefulness for prognosis. Since quantification of colocalization not only proves its existence objectively, but also provides many important additional details about it (dynamical changes, contribution to colocalization of a particular antigen/channel, etc.), it is natural to expect that the use of this methodology should significantly extend the significance of these and similar observations by making them less descriptive and much more factually-based. In addition to the peripheral proteins, it can be expected that changes of the degree of colocalization can have diagnostic and prognostic values when examined in the nuclei of tumor cells, such as for example colocalization of promyelocytic leukemia (PML) bodies with telomeres (Cancer Res 2007;67:7072).

It should be mentioned, however, that the use of quantification requires a certain level of methodological knowledge and skills to ensure that it is executed correctly. It is also critical to ensure that images for analysis are properly acquired and processed prior to quantification. Fortunately, this issue received a much required attention of late with the appearance of several high quality articles stressing the importance of accuracy and precision in quantitative fluorescence microscopy and even describing a specialized quantitative colocalization analysis protocol. An another important thing worth noting is that, to be actually useful for diagnostics, quantification of colocalization should be at least semi-automated. This can be addressed by employing the latest software tools when, for example, the whole workflow of image analysis is saved as a preset or even as an action (similar to the Photoshop action), which then can be re-used and re-applied easily and quickly on as many images as needed. The critical issue here is to ensure that all images are prepared for the analysis and then analyzed in exactly the same fashion so that the results of calculations are easily reproduceable and fully comparable.

To conclude, we will likely see a noticeable increase in the number of studies attempting to use quantification of colocalization in tumors for diagnostic and prognostic purposes in the near future. Lets hope they adhere to the methodological guidelines and report really promising and applicable observations.