Technical

The Process

Various yeast colors in our palette.  From left to right: black (violacein), gray (violacein), white (wild type yeast with our selective marker), blue (anemone gene), purple (amilCP from coral), pink (RFP), light pink (beta-carotene), magenta (jellyfish), oranges/ yellows (beta-carotene).  Yeast pigments curtesy of Leslie Mitchell, James Chuang, Jasmine Temple, and Michael Shen.

Various yeast colors in our palette.

From left to right: black (violacein), gray (violacein), white (wild type yeast with our selective marker), blue (anemone gene), purple (amilCP from coral), pink (RFP), light pink (beta-carotene), magenta (jellyfish), oranges/ yellows (beta-carotene).

Yeast pigments curtesy of Leslie Mitchell, James Chuang, Jasmine Temple, and Michael Shen.

New York City sunset drawn using a tablet, Jasmine Temple

New York City sunset drawn using a tablet, Jasmine Temple

Pigments

We have taken advantage of just a small fraction of the chromatic diversity exhibited by nature to create some of our pigment vectors. Using both yeast Golden Gate (yGG) cloning and the versatile genetic assembly system (VEGAS), we are able to easily assemble and combine transcriptional units (TUs) consisting of a defined promoter (usually ~500 bp upstream of a start codon), coding sequence, and terminator (usually ~200 bp downstream of a stop codon). yGG cloning has allowed us to screen pools of promoters and terminators in a combinatorial fashion, leading to empirical determination of sets resulting in the desired shade of a pigment.

For single gene pigments, we have successfully used a modified 2m shuttle plasmid with a marker conferring resistance to either G418, nourseothricin, or hygromycin B as the backbone for our pigment production. Pigment encoding genes have been taken from multiple sources, including Red Fluorescent Protein, the sea anemone Actinia equina, and the biosynthetic pathways for b-carotene and violacein. Antibiotic markers ensure that contaminants are unable to grow on our agar canvases (which are made with antibiotic).

This approach works equally well with fluorescent proteins (under fluorescent light).

 

Image breakdown and color matching

We have written two complementary Matlab scripts to generate printing instructions for the Echo liquid handling robot. Briefly, they:

  1. Rescale the image of interest for the dimensions of our agar trays

  2. Match each pixel to the “closest” color of yeast in our collection using a user-defined colorspace

  3. Choose whether or not to print whitespace

  4. Output a csv file containing information about source well, destination location, and transfer volume for each pixel

Printing

The CSV file is then given to the Labcyte Echo 550, which uses acoustics to shoot precise nanoliter droplets onto a rectangular agar plate in the pattern from the CSV file. The plates are 192 pixels by 128 pixels, so the image can be made up of up to 24,576 droplets!

The plate is left to grow at 30˚C until the pigments begin to emerge, then placed in a cold room to intensify! 

Two days at 30˚C

Two days at 30˚C

After another 14 days at 4˚C

After another 14 days at 4˚C