Our first main challenge of the summer was to determine which organelles we could easily and distinctly target in order to produce our MiCodes. We referenced the YeastGFP database of global analysis of protein localization studies in the budding yeast, S.cerevisiae, which was compiled by the O'Shea and Weissman labs at UCSF. This database consists of microscopy analysis and characterization of GFP tagged to every ORF in yeast.
We found proteins that localized to various organelles, PCR'd them off of the genome, then tagged them with a fluorescent protein. After analyzing their specific localizations through microscopy, we narrowed our choices hoping to find consistent morphology and geometry.
We chose four organelles from a list of over ten candidates based on the following criteria:
- Existing targeting proteins or signal sequences in the literature.
- Visible distinction from the other chosen organelles.
The following were our final four choices:
H2A2-Venus The localizing protein is H2A2/YBL003C, a H2 histone protein that is essential for chromatin assembly. C-terminal protein fusion to H2A2 will localize the complex to the nucleus.
mKate-CIIC The C-terminus CIIC signal sequence is derived from the Ras protein, a signal protein located in the plasma membrane that activates in response to extracellular signals for growth and differentiation.
ABP1-Venus Actin-binding protein of the cortical actin cytoskeleton, important for activation of the Arp2/3 complex that plays a key role actin in cytoskeleton organization; phosphorylation within its PRR (Proline-Rich Region), mediated by Cdc28p and Pho85p, protects Abp1p from proteolysis mediated by its own PEST sequences.
ZRC1-mTurquoise The localizing protein ZRC1/YMR243C is a vacuolar membrane zinc transporter. C-terminal protein fusion to ZRC1 will localize the complex to the vacuolar membrane.
*Protein characterizations provided by the YeastGFP database*
We used RFP, GFP, and CFP (which show up in red, green, and blue imaging channels, respectively) and our four organelles that could be easily distinguished: the nucleus, vacuolar membrane, plasma membrane, and actin. The size of our MiCode set can by calculated using the function (2^x)^y where x represents the number of imaging channels while y represents the number of organelles to which the fluorophores are targeted. With three imaging channels and four organelles, we can create 4,096 unique MiCode combinations!
Our test MiCode set can accommodate up to 4,096 members!
Although a MiCode set of up to 4,096 unique combinations is plenty for our proof of concept experiments, many experimental libraries would require a set size close to the magnitude of 10^6. With a few augmentations we can easily expand the available MiCode set to and even past 10^6 unique combinations.
Micode set size can be expanded by increasing the 3 criteria.
- States: A general binary system consists of genes expressing fluorescent proteins can either be on or off. Additionally, photosensitive fluorescent proteins that are activatable by UV light can be used as an extension into a ternary system in which the states are: on, off , or photoactivatable (or PA for short).
- Channels: We currently utilize three channels: red, green and blue. It is possible to use other fluorescent proteins that are viewable under other imaging channels without overlap.
- Organelles: We optimized the use of four organelles (actin, plasma membrane, nucleus and vacuolar membrane) for our project but additional organelles can be utilized.
The more generalized equation for calculating the possible set of MiCodes.
By simply adding one more organelle and one more fluorescent protein that can be visualized under an additional channel,
the available MiCode set size increases by 3 orders of magnitude!
When choosing promoters for our cassettes, we had to find promoters that would make each element of the MiCode visible under the microscope. Furthermore, the promoters would have to adjust for differences in protein abundances and consequently the relative brightness for each localization tag. A portion of our project was to decide on this optimum set of promoters.
Out of the 7000+ promoters in the yeast genome, we decided to experiment with 5 promoters: pTEF1, pTDH3, pRPL18B, pRNR2, and pREV1 - listed from strongest to weakest. These were chosen based on the data from a paper from UCSF (Ghaemmaghami, et. al, 2003) in which every yeast genome open reading frame was tagged with GFP and observed for protein abundance. The 5 promoters listed above were simply 5 promoters that spanned a reasonable range in promoter strength and were promoters that Dueber lab was already familiar with. The promoter library we created allowed us to systematically compare all 5 promoters. Each yeast strain in the library expressed only one organelle localization tag, with one promoter, and we imaged them all under the same microscope settings. Two modes of image analysis were used for thorough comparison: one with the pixel brightness range set from 0-4095, and another with range 25-1000. The data can be seen here.
To assay their potential use in our project, we characterized the strength of several registry yeast promoters. We wanted to compare their consistency of expression and relative fluorescence, then quantify them against a standard promoter fluorescence. The promoters we used in our assay:
- pSTE5, a weak constitutive promoter for the yeast scaffold protein STE5. Registry Part: BBa_I766557.
- pCYC1, a mid-level constitutive promoter for yeast Cytochrome c. Registry Part: BBa_I766555.
- pADH1, a strong constitutive promoter for yeast Alcohol Dehydrogenase. Registry Part: BBa_I766556.
- pTDH3, a strong constitutive promoter for the yeast glycerol synthesis protein GPD1. Registry Part: BBa_K124002.
Experimental Design: We designed our promoters to express a fluorescent protein so that we could quickly measure bulk fluorescence via TECAN. To test if downstream sequence affected expression, we cloned the promoter in front of two different fluorescent proteins, yellow fluorescent Venus and red fluorescent mKate in different strains. If expression was sequence-independent, we expected similar fluorescence intensity from both channels. The device was cloned on a backbone with Leu2 marker and Cen6 origin of replication and transformed into S228C S. cerevisiae. Ideally, we wanted the calibration curve to span several orders of magnitude and exhibit a linear relationship, which was true for all the promoters we characterized except for pCYC1.
We collected data for 8 samples per strain. the averaged data that we acquired is shown below:
Calibration curve of fluorescence intensity. Error bars are +/- one standard deviation.
|background||55 ± 10||36 ± 2|
|pSTE5||126 ± 10||216 ± 20|
|pCYC1||189 ± 30||928 ± 30|
|pADH1||1170 ± 90||4360 ± 100|
|pTDH3||22164 ± 700||62531 ± 2000|