Hi, I'm doing the transwell invasion assay using MDA-MB-231 cells. The problem is that ,when I take the image after staining the cells, I found that the cells is not distributed evenly in the membrane, the fields under microscope were not uniform, and some field was full of cells and some field had little cells, resulting in big error in cells counting.Has someone met this problem? And how do you count it? Is it ok that count from the center and form a cross? Thank you very much.
Hi Xue, I had similar problems with quantitating cells that migrated from the top of the transwell to the underside.
What I eventually came up with was to ablate the cells from the top using Q-tips wetted In serum free media followed by a quick wash with TBS.
I then used an acid phosphate assay, with detergents, to quantify the internal acid phosphatase of cells left using a standard curve of known number of cells and their acid phosphatase activity.
A Q-tip swipe of the top and bottom was used as a background control.
The acid phosphatase assay is described in
http://www.sciencedirect.com/science/article/pii/S0003269796903838
Dear Xue,
We usually take images of most of the transwell insert at low magnification using a confocal microscope, You can take the images with the transwell insert still inside the multiwell plate, or put it on a glass bottom dish in order to improve the quality of the images. With a good 4x lens, at 2048x2048 pixel resolution, we acquire a large 3x3 mm2 field, which covers most of the membrane. The images have enough resolution to detect individual nuclei (stained with DAPI, Hoechst or Propidium Iodide), and we find no need to do Z-stacks. Afterwards, we automate counting of nuclei in múltiple images using a macro created with FIJI/ImageJ. We find this method fast and simple, and we avoid the problem of uneven distribution of cells. Let me know if you need any additional information.
Hi Ricardo, do you take images of the whole transwell inserts, or do you ignore taking pictures of parts the inner membranes that are deemed to have too few cells?
How do you quantitate nuclei that are stuck in the pores? Do you count them as migrating cells?
ImageJ is great at counting cells if you have certain parameters of cell size, but if you are just counting nuclei, measurement of cell migration in transwells is really subjective.
Plus ImageJ can't quantify cells that have migrated to the sides or the bottom of the transwell inserts
I forgot to mention that, before imaging, we remove the top part of the insert with a cotton swab, as it is usually done. The single plane image acquired with a 4x lens in our confocal microscope is about 30-50 microns in axial resolution (very poor), but this is good to obtain most of the cells in the membrane and eliminate the rest. Yes, I agree that it counts cells stuck in the pores and at the inner part of the membrane, but it will be the same for all conditions. I attach an example image so you get an idea.
A simple ImageJ macro to count nuclei for this image would be as follows (threshold and size may need to be adjusted):
run("Median...", "radius=2");
setAutoThreshold("Default dark");
//run("Threshold...");
setThreshold(150, 4089);
setOption("BlackBackground", false);
run("Convert to Mask");
run("Fill Holes");
run("Watershed");
run("Analyze Particles...", "size=10-800 show=Outlines display exclude clear summarize add");
Hi Ricardo, I just want to understand your procedure.
Please correct me if I am wrong. After the migration, you swab away the top cells that haven't migrated and you stain the nuclei, cut the insert from the well and mount it on a slide and image it
I looked at your picture, but I could only see white dots on a black background and no perforations that cells migrated through or are stuck in.
Hi Steingrimur
After migration, we remove matrigel and top cells with a cotton swab, then fix and stain with Hoechst. Afterwards, we do not cut the insert and mount it on a slide, which is the tedious part (the filter does usually not remain flat). We simply remove the whole insert from the well, put in on a glass bottom petri dish and take images with a confocal microscope, You can also take images with the inserts still in the multiwell plate, which is faster, and the images are still good enough to perform the quantification analysis.
Regarding your second question, yes, the porous membrane is there, but we only capture the fluorescence image. You can capture the transmited light image easily, but we find no need in doing it. We also have scanning electron microscope images of this same experiments, where you can see cells going through the pores. I could upload an image if you want it.
Hope it helps
Hi Ricardo,
Thank you for your reply. Do you mean that acquire a large 3x3 mm2 field, which covers most of the membrane and the transwell insert diameter of 24 well is about 3.2mm? So maybe the best way for me is to count all the cell in the membrane.This is the most accurate way to count.
Hi Xue Sun,
Have you looked at the technology offered by ACEA Biosciences? They offer devices that monitor invasion real time, providing highly accurate quantitation using an impedance-based principle. This eliminates chances of error and is easy to use. The company also provides free support and offer demos.
https://www.aceabio.com/products/rtca-dp/
The Transwell method (as you may have noticed) is very tedious and very error-prone for beginners.
- Badges
- Science method