Lab skills quiz – Microscopy

Phase contrast microscopy doesn’t allow you to see fluorescence

electron microscopy allows you to obtain high-resolution structural information, for this hypothesis you would image the two cell types and compare the “inner plaque” of the hemidesmosome.

TEM isn’t used for stiffness measurements, perhaps atomic force microscopy could allow you to test this hypothesis.

For topography, you would be better going for a scanning electron microscope approach

For this hypothesis you would measure RNA levels, most likely by qRT-PCR. If you were desperate to use a microscopy technique, then in situ hybridisation would be an option but this could be difficult to do quantitatively in this context.

Although you could quantify staining intensity, this sort of technique is actually really bad for this type of quantification – no clear linear range, lots of bleaching, variability in processing; lots of people do this but it’s really unreliable and should be avoided! If you want to quantify proteins, use an ELISA, a western blot or some other quantification method

Location yes, quantification not very well! Although you could quantify staining intensity, this sort of technique is actually really bad for this type of quantification – no clear linear range, lots of bleaching, variability in processing; lots of people do this but it’s really unreliable and should be avoided! If you want to quantify proteins, use an ELISA, a western blot or some other quantification method

You would need fixed cells and tagged proteins for this sort of analysis.

shorter wavelengths carry more energy that longer wavelengths. Although white light contains short and long wavelengths, once you account for shorter exposure times the damage is less. If you are planning a live imaging experiment, you should take this into account when deciding how long/how frequently to image.

With this set up you could image sequentially: for blue, excite with 405nn, BP380-450 will cut out any green light, for green excite with 488nm, BP400-550, although you will get a little bit of red excitation with the laser, you will cut out the red emission with the filter. red excite 543nm BP500-600 will excite a little green/purple but the band pass cuts out the emission from those fluorophores, purple excite 650nm, LP700 again this will cut the red emission.

it is possible! Try again.

Using this scope you are relying on the filters to cut out the emission from the wrong fluorophore. the BP380-480 will be OK for blue, but the BP400-580 will not only get green light but also some blue and some red. i.e. even when you want to look at red, you can’t be sure you don’t also have green. Note shorter range filters exist, these numbers are just made up! How you will image your samples is something you should be considering during your design stage of these experiments.

this won’t work for you as you won’t be able to excite the purple fluorophore

there is a decent amount of overlap between the emission curve of green and the excitation curve of red. There are better FRET pairs than this, but it could work.

FRET only works if the donor is shorter wavelength than the acceptor, remember the emitted light is longer than the excitation (Stokes’ shift)

there is very little overlap here, you might get a little FRET but it wouldn’t be very good.

 there is very little overlap here, you might get a little FRET but it wouldn’t be very good.

there is no overlap here, you wouldn’t get any FRET

Nope… what is the feature of a lens that determines the resolution? This objective is actually an extra long working distance, you would use this for imaging through thick substrates but often you have to compromise in terms of resolution or sensitivity.

Yes, numerical aperture determines the cone of light that enters the lens, the higher the better.

Try again.

No, not this one… have another go.

Yes! Learn how to adjust this to make sure you get even illumination for your phase contrast optics

adjusting the stage will change the focus. Here the sample is in focus across the image, the issue is how the light is illuminating it

this is unlikely to be where the issue is here, usually if there is an issue with the way the objectives are mounted you will have focus issues or no image

You will close the diaphragm to help you fix the problem, however this isn’t the part you will adjust

Yes, this is why your samples will fade as you image them. This feature of fluorophores and fluorescent proteins can be used to your advantage: in the techniques FRAP and FLIP. FRAP – fluorescence recovery after photobleaching – can be used to measure the rate at which turnover occurs within a complex. FLIP – flourescence loss in photobleaching – can be used to ask similar questions but usually is more focused on whether there are distinct populations of protein e.g. pop 1 assembled into an adhesion complex and pop 2 in the cytoplasm

Yes, the information is in the figure legend however, the reader shouldn’t have to read the legend to understand the figure and simply improving the labelling would help

Yes, seems obvious but this sort of sloppiness takes seconds to fix so why choose for it not to be right?

I agree with you, however, this is a style decision by the journal rather than the author! When the figure was prepared it would have been on a white background!

If that helps they should be there but it isn’t really necessary

These type of images are typically taken using a colour camera, it is usual to show them in the colour they were captured in. The white balance of the camera could perhaps have been adjusted but the colour scheme is OK.

At present, superresolution microscopy requires fluorophores that “blink”, a GFP conjugate will give a steady signal and so we won’t gain resolution here compared with antibody approaches

in live imaging you are using a protein to generate your signal, this protein will degrade over time and therefore will fade. Antibody stains will fade too but there are a number of anti-fade reagents that mean that processed slides can kept for longer without dramatic reductions in signal. Simple message; if you are doing tagged protein work, make sure to image as soon as possible

this is always something to have in mind, we’ll have a look at how to control for this issue in the next question.

No, you can fix your fluorescent protein expressing cells and then image them afterward.

No, you could introduce another fluorescent protein eg with red (tdTomato, mCherry), yellow (YFP), blue (BFP) or other tags. The entire visible spectrum as well as far reds are covered. You could also combine fluorescent protein with dyes and antibody staining as required.

Yes, this would increase your confidence that it isn’t a paxillin specific effect but rather more generally associated with focal contact turnover. There is another correct answer in this list, have you identified it already?

You almost certainly should run this control. It allows you to see the effect of the GFP on the cells and you would infer from this that the difference in effect between this and the paxillin GFP is due to the paxillin. However,  this wouldn’t help you know if the paxillin is assembled into focal contacts or some other cellular structure. Remember to design your controls to match your question

Increasing experimental “n” will increase your confidence about the generalisability of your finding. You will need more samples to generate the numbers required for statistical interpretation. However, increasing n won’t in itself add depth to the experimental interpretation, rather you will be confident that what you have observed is real (this is important too!)

This would allow you to confirm that the GFP signal you have been tracking is (at least at the end of the experiment) associated with focal contacts. It’s not perfect by any means, but it would be quick and simple to do and would increase the confidence in your results interpretation. There is another correct answer in this list, have you identified it already?

Well, I didn’t really expect this one to get picked! If the cells are removed, the signal that is left will be static and just fade over time. It won’t help us here. Try again.

This isn’t fair. What is the reason for the difference between the old data and your current data; likely that reason can account for the difference in your treatment images. You are misrepresenting your data. Don’t do this!

This isn’t fair ; your images should tell the story of the population, if there is variability, then that should be part of the story. Your quantification of your images will tell you what is truly representative.

 As long as they don’t obscure or eliminate information then simple manipulation is OK

What is the point of the controls? If you mix and match experiments in your figure you lose the point of including them. Differences will exist between experimental runs that could account for why your data is what it is.