Methods
Things we do in the lab
Laser-Scanning Confocal microscopy
Confocal microscopy is our most used and preferred technique. In combination with fluorescently labeled proteins, we use confocal microscopy to look at the localization of those proteins inside muscles. Also, we combine it with specific mutations to see the structural consequences of those mutations.
Transmission electron microscopy (TEM)
Transmission electron microscopy and muscle research have a long history. The first TEM image of a Drosophila indirect flight muscle was done in 1963 (link). The method and the resolutions have not changed much but we still use it because it is a powerful and simple technique that allows seeing the sarcomeres at great resolution.
Flight assays
We primarily work on the flight muscles and a good way of testing their function is by monitoring flight (flies with dysfuctional flight muscles can't fly). Drosophila flies fly by moving their wings ~200 times per second. To see this movement we use high-speed cameras and lasers. The overall setup for monitoring flight is depicted in the figure above. A fly is placed between an infrared laser and a detector, the changes in light indicate the movement of the wings. We use this simple method to characterize flight problems in mutant flies. The figure above shows the difference between a normal fly, that can fly for more than 30 seconds, and a fly with a mutation in the filamin gene, that can fly but only for a few seconds.
Super-resolution microscopy
We use some super resolution microscopy techniques to see the muscle ultrastructure at high resolution. For fixed tissue, we use expansion microscopy and for live imaging, we use Super-Resolution Radial Fluctuations (SRRF).
Genetic manipulation of flies
The ability to make precise mutations and to integrate any DNA sequence into the fly genome makes Drosophila a great model system. We use CRISPR-Cas9 to target any gene, we use the Phi integrase system to incorporate DNA at a specific location, and we add fluorescent proteins to see the proteins and to make the flies glow. Here is a great historical review of common techniques used to manipulate the fly genome (link). Here is a review of CRISPR-Cas9 tools in Drosophila (link).
Deep Learning approaches to microscopy imaging
We are starting to develop deep learning approaches to compare microscopy images from control and mutant muscles. We mainly use the Google Colab notebooks from the ZeroCostDL4Mic project. Because the sarcomeres have very stereotypic shapes, deep learning approaches work well at predicting the position and shape of their components. In the figure above, we trained a model to draw Z-discs based on confocal images of actin-stained sarcomeres (input image). Note that the predicted Z-discs and the real ones match almost perfectly.