One of the major challenges in using 3D cell culture models is fully utilizing their three-dimensionality and improved mimicry of the in vivo micro environment. While there are one dimensional dissolution based assays for characterizing these models, these assays are limited in the quality of data they can collect and thus researchers are highly interested in image based end points. However, the major problem with using image based endpoints with 3D cell culture models is characterizing all of the cells throughout the depth of these models.
If you were to look at any recent publication on 3D cell culture models that uses image-based end points you will see that optical Z slices look like donuts where the center of the Z slice is black and only the cells in the periphery of the model are characterized. Often, confocal Z stacks are used to illustrate 3D cell culture models, but these depictions miss-represent the data as this inherent bias is obscured. It has long been thought that this bias was due to the inability of labels (e.g. chemical dyes, antibodies) to penetrate into 3D cell culture models and thus it was only possible to characterize peripheral cells.
At Visikol, we are focused on the clearing, labeling and imaging of tissues and have shown an ability to label and image whole mouse brains with imaging depths as thick as 6 mm. Therefore, when we first heard about the inability to uniformly label 3D cell culture models, we were intrigued as these models typically have a diameter of less than 500 µm. It was our suspicion that this limitation was due not to the inability to uniformly label these models, but instead a result of optical attenuation. 3D cell culture models tend to be dense clusters of cells and light attenuates significantly, limiting effective imaging depth to approx. 40 to 70 µm.
To evaluate the cause of this imaging limitation, we labeled a 250 µm NCI-H2170 spheroid with a SYTOX green nuclear stain and imaged it using both confocal and wide field microscopy. We then took this same spheroid and cleared it with Visikol® HISTO-M™ and re-imaged it using both confocal and wide field microscopy. Through the application of clearing we demonstrated that the amount of cells characterized in the 3D model increased 3-4 fold across both imaging techniques. We also showed that the interior of the model was able to be characterized in its entirety following the application of tissue clearing.
Therefore, we were able to show that the characterization of whole 3D cell culture models is generally not limited by label penetration but instead optical attenuation which can be mitigated with a tissue clearing reagent. However, it must be noted that labeling time and concentration needs to be optimized for each individual label as some will take longer and higher concentrations than others to achieve uniform whole model labeling.