My task is to quantify double positive cells in the retina. Currently I use and antibody against BrdU (Alexa 555) and a transgenic reporter gfap:nls-gfp. Unfortunately the transgenic reporter is not evenly expressed so double positive cells might be missed, thus I am searching for a different way of acquiring images.
My current setup was to take tile scans at the confocal microscope (LSM780) with a 25x water immersion objective to obtain proper cellular resolution. The Scans also include Z-stack, since the sections are 14µm.
Problem here is that first the scans take very long approximately 1,5h and I have lots of samples to quantify. Another problem is that sometimes sections are not perfectly flat on the slides (e.g. air bubble below), which increases the Z axis and thus again the time component. Also the transgenic background is not perfect.
Regarding the time there might be also an issue of bleaching.
I was thinking about an enzymatic staining to eliminate some of the issues above but not sure if that will help, especially with colocalization. Another idea was to use Qdot antibodies.
Hope someone of you can help me with this.
You can average the ten high power fields (HPF) to count positive cells and comparing
Q dots are NOT antibodies but probes...I suggest you check out Life Technologies at
http://www.lifetechnologies.com/uk/en/home/brands/molecular-probes/key-molecular-probes-products/qdot.html
I think they have what you need
Not sure how you're mounting slides, but if you're doing fixed frozen sections you can place a very thin layer of normal saline on the slide before mounting. This should solve your bubble problem.
gfap:nls-gfp might be an overly complex way of doing this experiment. Can you simply stain GFAP and co-localize with mitotic BrdU cells? GFAP antibodies are some of the most consistent. The signal should be stronger- decreasing exposure time and reducing bleaching.
Hi Johannes,
I am not sure what microscopes you have access to besides the confocal. If you have access to slide scanners like the one we have in our microscopy core facility (www.ucalgary.ca/runcore check for the Slide scanner on the website), you can obtain images of the whole section in a very short time. The slide scanner also has the capability of adjusting the Z focus by building a focus map and uses that Z plane when it images that area. You can also obtain a Z stack of the section and use the Extended Focus Image (similar to Maximum intensity projection) for the whole section. You can try wide field fluorescence microscope instead of confocal which captures more light than a confocal. Make sure you have a good deconvolution or deblurring post-processing software.
Regarding the quantification: Are you performing an exhaustive count on your sections (ie. counting each and every cell in each section). Have you thought about stereology method of quantification as it is more efficient and provides a good estimate of the cell number. If the cells are not distributed uniformly, then the HPF method suggested by Rong-Kung will not work as the cell count will be biased where you are counting. The cell count is also affected by the reference volume/ area and the amount of shrinkage the tissue undergoes (uniform/ non-uniform shrinkage) between the experimental groups and cause variability in the count and they call it the "reference trap". If you need more information about stereology, please contact me at [email protected] and we can discuss how it can be solved or your experiment can be designed. Or post your question in The Stereology and Microscopy Info group in LinkedIn where we have many members who are experts in the field of stereology. Good luck.
Thanks for the help so far.
Jacob D Lescher
I am mounting with 70%Glycerol/30%PBS. When placing saline on top of the frozen sections, do I have to leave it there or just incubate shortly? and you mean saline as a solid substance?
about the gfap staining I would prefer the nuclear signal since this allows a more precise way of colocalization and I am not aware of any antibody staining müller glia in the nucleus.
Sathya Srinivasan
For now I simplified the experiment by using a Crosshair for colocalization.
Nonetheless I am going to ask our imaging facility about slide scanners and/or the wide field microscope.
Actually I also think that stereology might give a more accurate picture of the actual cell numbers.
I have a few tips to add, if it is not too late:
1) Sometimes the "optimal" distance between optical slices when taking a Z-stack is much finer than what is needed (i.e. you end up with redundant information). Unless you plan to do orthogonal optical slices through your Z-stack afterwards, sometimes it is unnecessary to obtain such high resolution. Reducing the number of images you are obtaining in a Z-stack (same range, images more spaced in Z plane) can substantially decrease the amount of time is takes to acquire each one without having much effect on your projected image.
2) To get rid of bubbles when I was longitudinally sectioning the optic nerve, I would use a very soft, small paint brush to put a drop of PBS over the section, after a few moments the section would loosen slightly and the bubble could be maneuvered out.
3) Many years ago I had difficulty visualizing GFP expression in a transgenic fly line, and had very good luck using a GFP antibody with an Alexa 594 secondary (could also try a far red secondary). Might work well, but also adds time. I agree with Jacob Lescher that GFAP antibodies are very strong and easy to use.
All the best
- Badges
- Science topic