Penetration Depth in 3D Immunostaining（Part I)

Penetration of 3D immunostaining

Penetration is critical to achieve reliable data analysis. It encompasses both the penetration depth and uniformity.

Penetration depth refers to how deeply specific staining signals can be seen from the nearest tissue surface. It can be quantified as the distance from a given signal to its nearest tissue surface. A signal can be seen only when it is significantly above the background, usually by two standard deviations.

Uniformity refers to how the signal-to-background stays constant across penetration depth. If you have two cells expressing equal amounts of a marker protein, at varying distances from the tissue surface, they should have nearly identical signal-to-background ratios. 

How to know when there is a penetration problem

Gradients of signals that are much stronger at the tissue surface but weaker at the tissue core are suggestive of penetration problems. Unless there is a strong biological reason to support such a surface-strong core-weak signal pattern, it is undesirable for a 3D IHC image to exhibit such gradients, as this can lead to severe quantification biases.

Penetration problems are most obvious by scrolling through the image stack; gradients along the tissue surface will appear as a bright rim or border completely surrounding the tissue contour. In publications, sometimes the authors might hide penetration problems by presenting pretty 3D renderings and not presenting a cut-through view. In that case, you can look for a bright “shell” with “empty core”, where features seem to be most obvious on the surface but not within. Sometimes, this can indeed be hard to spot on a static figure.

To distinguish between biological non-uniformity and penetration non-uniformity, the easiest way is to stain your protein of interest in 3D, cut it in the middle, re-stain for the same marker with another fluorophore, and image the cut surface (see the schematic below). The pre-cut 3D staining should correlate well with the post-cut 2D staining. This is because staining after the cut should be devoid of both probe and light penetration problems, which represents the biological ground truth. 

How do we quantify penetration depth and uniformity?

You can quantitatively compare the penetration depth of various techniques using a peer-reviewed method (https://www.sciencedirect.com/science/article/pii/S2667237523000772). The experiment is essentially the same as described above, but now the pre-cut 3D-staining signal is divided by the post-cut 2D-staining signal to obtain a ratio. The ratio data is then plotted against their penetration depths. In the ideal case, the scatter data points should hover around a flat line, indicating that 3D staining reveals signal intensity variations just as good as 2D staining, independent of penetration depth. In most cases, penetration problems will show up as an exponential decay curve, and the obtained decay constants can be quantitatively compared to one another.

A code is available to illustrate such a comparison (https://codeocean.com/capsule/1935786/tree/v1).