Summary: Visikol HISTO-M showed clear superiority for imaging immunolabeled 3D cell culture models when compared to ScaleS4 and CUBIC tissue clearing approaches. Visikol HISTO-M consistently showed a much higher signal to background ratio than CUBIC or Scale.
Clearing tissues for microscopy is a developing field with a growing list of options. There is no one size fits all solution available and making the right choice depends on the type of tissue sample being used as well as how it is being labeled. We have a table summarizing the various clearing systems available that describes these specific advantages and disadvantages.
However, the focus of this post is on Scale and CUBIC in as there have been several papers recently describing their use for multiple applications. Scale and CUBIC are both aqueous clearing methods that use the hyperhydration and delipidation from urea to clear tissue samples. Unfortunately, they do not integrate easily with immunolabeled tissue. To illustrate this problem, we labeled HepG2 spheroids with a nuclear stain and a Pan Cytokeratin (Pan CK) label, followed by clearing with Visikol HISTO-M, Scale S4, and CUBIC.
A 96 well Corning ULA plate was used to generate HepG2 spheroids. After one week of culturing the spheroids were identically fixed, permeabilized, and labeled with primary 1:200 mouse Pan CK and secondary 1:200 anti-mouse 647 with 1:1000 DAPI. Four spheroids were cleared for each condition. For Visikol HISTO-M, the spheroids were dehydrated through a Methanol gradient and finally exchanged into Histo-M. For Scale S4 the PBS was exchanged three times with Scale. For Cubic, the spheroids were exchanged into Cubic 1 for 24 hours and then into Cubic 2.
The 96 well plate was left in the refrigerator for 2 weeks and then imaged on the 405 and 647 channels every 10 seconds for 1 hour. To process the data, the spheroid ROI was identified using automated thresholding on the first frame. The background ROI was manually defined by selecting a clear round area of similar size to the spheroid that did not contain any debris. The raw intensity per area for each ROI was then calculated and graphed.
The DAPI channel gave similar results with strong signal to background from each of the clearing solutions, exhibiting signal to noise ratios of approximately 30 for each clearing agent. Examination of the antibody signal showed a significant advantage of using Visikol HISTO-M with immunolabeling. Because Scale and CUBIC rely on urea for clearing, which denatures proteins, the secondary antibodies lose their affinity to the target and are pulled off the tissue and into the solution creating very high background signal. The low resulting signal to background makes automated thresholding a challenge and can add additional work to the data analysis stage of image processing.