Embedded Files
My research is aimed at understanding how proteins control membrane structure and dynamics, specifically during the process of autophagy. At the current time there are three projects running the lab
Mapping the interaction sites of Atg11, a key protein for selective autophagy, with its protein partners in the Autophagy Initiation Complex.
Using yeast genetics and electron microscopy to determine which autophagy proteins affect autophagosome size and which affect autophagosome number.
Improving a computer simulation used to estimate the 3D size and number of autophagic bodies from the random 2D slices generated by electron microscopy.
We've also recently trained and published a cellpose 2.0 model for automated labeling of autophagic bodies in electron microscopy images of yeast vacuoles.
If you would like to learn more about these projects, keep reading, or click here for a general background on autophagy.
Atg11: Organizing Selective Autophagy
Atg11: Organizing Selective Autophagy
Atg11 is a large protein with predicted coiled-coil domains (CC2, CC3 & CC4) that is a key component of the Autophagy Initiation Complex during Selective Autophagy. It is required for selective autophagy in yeast, and has been shown to interact with a number of other proteins in that complex, such as Atg1 and Atg9. The regions of Atg11 important for those interactions have been roughly mapped via deletional analysis by various labs to the CC2 and CC3 domains, but not to individual residues, nor do we know how these partners are organized in space or time. Interestingly, 6 different proteins partners have been reported to require CC2 for their interaction, even though this domain is only 40 amino acids long. We performed systematic mutational analysis of the CC2 and found three residues that were critical to the binding of all of these partners, perhaps because they play an essential role in the overal structure of Atg11 (Meyer et al. 2022). We are continuing to search in other regions of Atg11 for residues that are specific to the binding of individual partners.
Autophagosome Size and Number
Autophagosome Size and Number
When yeast cells are starved, they begin generating large autophagosomes that wrap up cytoplasm and then fuse with the vacuole to deliver their contents for digestion. If we make yeast mutants that cannot digest those contents, then they accumulate in the vacuole, where they are termed "autophagic bodies," and can be imaged by electron microscopy. We can measure the size and the number of the autophagic bodies and use this to help understand the roles of the different Atg proteins by investigating what effect different levels of these proteins have.
For example, previous research in the Klionsky lab showed that Atg8 controls autophagosome size (less Atg8 = smaller autophagic bodies)1 , whereas Atg9 controls not size but autophagosome number (less Atg9 = fewer autophagic bodies)2. Recently published research from our lab has shown that Atg7, the first enzyme in the Atg8-conjugation cascade, controls both size and number (less Atg7 = smaller and fewer autophagic bodies)3 . This is surprising, since the only known role of the Atg7 pathway is the formation of Atg8-PE, and Atg8 only controls autophagosome number, not autophagosome size. This suggests that Atg7 may have an additional role, perhaps related to its formation of intermediate members of the cascade such as the Atg12-Atg5 conjugate.
Currently we are following up on these findings by examining the effects of Atg3 and Atg10 on autophagosome size and number, as well as exploring other components of the autophagosome formation machinery.
1. Xie et al. Mol Biol Cell. 2008;19:3290
Top: Electron microscopy image of a crossection of a yeast cell showing autophagic bodies in the vacuole. The average size and number of these bodies can tell us something about the role of the Atg protein that was manipulated in this cell.
Bottom: Atg7 catalyzes the first step in the enzymatic pathway that attaches Atg8 to the lipid PE. Atg8 only controls autophagosome size, but Atg7 controls both size and number, suggesting that there may be additional roles for some of the intermediate components, such as perhaps the Atg12-5 conjugate. Diagram by Jacquelyn Roberts, from her EMU Honors Thesis.
Autophagic Vacuole Simulation
Autophagic Vacuole Simulation
Transmission electron microscopy images capture a thin, essentially 2D slice through the autophagic bodies. A 2D slice through a 3D object doesn't tell you everything about that object, but since we have many hundreds of random slices through different bodies, we can use the distribution of this data to estimate the original distribution of the size and number of the 3D bodies in the cell.
There are published methods to do this estimation, which have proven useful in comparing the effects of some mutations on autophagic body size and number1. However, a limitation of the accuracy of these methods is that they model the autophagic bodies as perfect spheres, whereas we can tell from looking at the electron micrographs that they are not perfect spheres, but instead smoosh together into irregular shapes.
One current project in my lab is focused on improving these simulations so that they can model the autophagic bodies as realistic smooshed shapes. To do that we are using a free pixel-based cell-modeling software called "CompuCell3D," and writing python code to integrate this into the rest of the "Autophagic Vacuole Simulation" pipeline.
The most current working version of the simulation software (which still models the bodies as perfect spheres) runs in R. It is available on my Github page, "Autophagic-Body-Number-Simulation."
Top: Electron micrograph of a starved yeast cell, with autophagic body profiles outlined in yellow. Note that the bodies are not perfectly round, but instead smoosh together.
Bottom: Results of a test run from a CompuCell3D simulation that started with perfect spherical bodies and then smooshed them together into more realistic shapes.
Page updated
Report abuse