To verify whether or not the fluorescent proteins we attached to Sup35’s prion domain still fluoresce, we expressed them under the Cup1 promoter in the W303 strain of yeast. We had variants of the strain that were both [PSI+] and [psi-] (ie. strains that had aggregated and soluble Sup35 respectively), and expressed our constructs in both.

As stated in our section on Cup1 promoter characterization, we observed that empty yeast cells never had more than 0.2% of their population expressing a fluorescence greater than 1000 fluorescence units as measured by our flow cytometer. Therefore, we defined “fluorescent” cells as cells having a fluorescence of greater than 1000 fluorescence units.

We observed yeast samples consisting of 10,000 cells in a flow cytometer that used a 488 nm laser. This laser is capable of moderately inefficient excitation of YFP and very inefficient excitation of CFP. Monitoring of fluorescence was performed with a channel that accepted any emission from 505 nm to 560 nm.

The data we collected are formatted as follows. Y is used as a shorthand for our construct PrD-YFP, and C is used as a shorthand for our construct PrD-CFP. The samples are identified by which constructs they contain. There were two separately performed transformations of PrD-YFP into [psi-] yeast to verify that that particular part is fully functional. [psi-] YFP R (“R” standing for replicate here) was done separately, and all other samples were transformed together on the same day. Average fluorescence is provided in fluorescence units as measured by the flow cytometer. Additionally, the average fluorescence provided is the average of the fluorescent population ONLY.

The presence of our constructs in our yeast was confirmed by nutritional selection during transformation, and replica plating of samples after transformation onto nutritionally selective media. All transformations were efficient (100s to 1000s of colonies), indicating no substantial problem.

Samples were taken directly from transformation plates and suspended in phosphate buffered solution (PBS), as this was the buffer used in the flow cytometer. The reason why samples were taken straight from the plate was because fluorescence decreased rapidly once samples were inoculated in YPD (though it persisted at a low level in most samples after 12 hours).

Replicate 1:

Replicate 2:

Any data point with fewer than 0.2% of its cells fluorescing was deemed non-fluorescent as per the explanation above, and this is why average fluorescence values of such samples were not calculated.

There is fluorescence in all [psi-] yeast barring one sample, and there was no fluorescence in the [PSI+] samples, baring the two [PSI+] Y.

In order to determine if [psi-] C was fluorescing because of our construct, we checked its fluorescence value after 12 hours of growth in YPD. A significant part of the population demonstrated fluorescence (0.3% of its population). Therefore, we concluded that [psti-] C was indeed fluorescing in the flow cytometer, but very weakly due to the suboptimal wavelength of the laser used.

From this data, we conclude that (1) all our parts are functional in that they fluoresce, and (2) that all our parts are functional in that they interact with the Sup35 aggregate, as was intended. (1) is supported by the fact that fluorescence was consistently observed in the [psi-] state, and (2) is supported by the fact that for all samples, fluorescence was either reduced or eliminated in [PSI+]. Since the presence of the Sup35 aggregate was the only difference between [psi-] and [PSI+] cells, we can deduce that it must have caused the decrease in fluorescence.

The finding that fluorescence decreased when our fusion proteins joined the aggregate is actually quite interesting. It suggests that our selection of the prion domain used as a tag may have been suboptimal. We chose to tag our fluorescent proteins with the first 137 amino acids of Sup35, since this is the region required for the maintenance of the Sup35 aggregation behaviour [1]. However, the inclusion of the entirety of the N and M domains of Sup35 (which make up the first 256 amino acids) may have been more appropriate, because other labs have successfully tagged GFP with this section of Sup35 and observed it joining aggregates and fluorescing in them [2]. These extra amino acids could have created more space between the amyloid core and our fluorescent proteins. As is, the proteins may be getting buried in the aggregate, or they may be joining the aggregate before they can fold correctly.

References:

[1] Bradley, M. E., & Liebman, S. W. (2004). The Sup35 domains required for maintenance of weak, strong or undifferentiated yeast [PSI+] prions. Molecular Microbiology, 51(6), 1649–1659. https://doi.org/10.1111/j.1365-2958.2003.03955.x

[2] Tyedmers, J., Treusch, S., Dong, J., McCaffery, J. M., Bevis, B., & Lindquist, S. (2010). Prion induction involves an ancient system for the sequestration of aggregated proteins and heritable changes in prion fragmentation. Proceedings of the National Academy of Sciences, 107(19), 8633–8638. https://doi.org/10.1073/pnas.1003895107