Here's how to play Colisweeper!

Figure 1: Project description

Inspired by the computer game Minesweeper, we aspire to incorporate all the same features and rules in the biological version. To replicate the "mouse click" in the computer game, you can add a substrate to the colony of your choice which leads to a colorful enzyme-susbtrate reaction. The agar mine field consists of colonies which can be one of the following:

a non-mine colony, which will display the "safe colour" when played
a non-mine colony located close to one or more mines, which on addition of a multi-substrate reveals the number of mines in the vicinity of this colony
a mine colony, which when played reveals the mine color and the game is over

You also have the option to flag a mine - if you are certain that a colony is a mine, you can flag it using a second substrate.

The biology behind the game

Figure 2: Signal transduction from secreted signaling molecule to colorimetric response.

Each minesweeper square in the computer game corresponds to a bacterial colony on the agar mine-field. We have two bacterial strains: 1. the mine strain with the sender cells and 2. the non-mine strain with the receiver cells. The cells communicate through diffusion of OHHL and change color to give the player information in order to logically carry out the next move in the game. The colonies remain white as seen on the agar plate until any substrate is added. The left and the right click of the mouse is simulated with the addition of the multi-substrate mix and the single substrate.

The biology is explained here. The sender colony secretes the quorum sensing molecule 3-oxo-N-hexanoyl-L-homoserine lactone (OHHL) that diffuses through the agar to the surrounding cells. The receiver cells passively take up the OHHL by diffusion where the signaling molecule forms a complex with the inactive LuxR to activate it. The information is translated via mutated pLuxR promoters of different OHHL affinities which leads to the secretion of different hydrolases. Within minutes after the addition of substrate a change in color indicates the identity of the played colony and number of surrounding mine colonies.

The Game Board

Figure 3: The game board pattern

The traditional minesweeper pattern is a grid with tightly spaced squares, where each square is surrounded by eight other squares. To replicate such a grid in the biological game is not feasible. Our mine colonies send the OHHL signal through diffusion in the agar (OHHL) and therefore must be placed in such a manner that equal amounts of signaling molecules are received in adjacent non-mine colonies.

If we located bacterial colonies in the place of squares like in the traditional game, the colonies placed in the corners would be further away compared to the colonies along the edges. We decided on a hexagonal honeycomb-like pattern, where colonies are placed at all edges of the hexagon except from the center. In this set-up the number of mine colonies restricted is to only one, two or three that can surround a non-mine colony. This also facilitates the selection of suitable reporter enzymes. To plate the field, liquid cultures of mine and non-mine cells are grown to an OD₆₀₀ of 0.5, then, using a pipette, 1.5μl of liquid culture are placed according to the grid.

What about signaling and information processing ?

Signal expression by mine cells (sender cells)

Mine colonies constitutively express the "mine-cell reporter” that reacts with one compund of the multi-substrate solution to produce a chromogenic product that indicates a mine colony. Addition of the multi-substrate turns the colony blue, indicating the mine and hence the game ends. At the same time, the mine colony is a sender module of OHHL. LuxI is expressed constitutevely and synthesizes the OHHL signaling molecules. This means that the mine continuously produces OHHL, allowing it to gradually diffuse radially out to neighboring colonies. The signaling molecule OHHL is processed by the receiver cells in the agar mine grid.

Information processing by non-mine cells (receiver cells)

Figure 4: [OHHL] dependent information processing

So what exactly happens with the expressed OHHL?

1. Receiver cells respond to low OHHL levels: If one mine cell is close to a non-mine colony and by OHHL diffusion takes place to reach the non-mine colony. Due to a highpass filter using the LuxR system in the non-mine colony,low concentrations of OHHL are detected and activate a reporter. Addition of the multi-substrate to this colony produces a salmon color, which indicates one mine is adjacent to the colony just played.
2. Receiver cells respond to high OHHL levels:If there are two or more mine colonies adjacent to a non-mine colony, a higher concentration of diffused OHHL will be processed by the receiver colony. The LuxR system works as a highpass filter again and induces the expression of another reporter to show that two or more mines surround the selected colony. Addition of the multi-substrate to this colony produces a magenta color that indicates that more than one mine is adjacent to this colony.
3. Receiver cells signal with no mines in their vicinity. If there is no mine adjacent to a non-mine colony, no OHHL is processed. Only the constitutive hydrolase is expressed, which on addition of the multi-substrate will give a yellow color. This indicates that there are no mines adjacent to this colony and the game can be continued.

Flagging
Remember what the flagging does in a traditional minesweeper game? Placing a flag on a mine prevents the mine to detonate. Flagging is possible in Colisweeper too. If the player is sure about a colony to be a mine, a second single substrate can be added on to this colony. Because all colonies express a constitutive reporter for flagging, a green colored product indicates a flagged colony. But remember, flagging does not reveal any information about mines in neighboring cells.

We thank our sponsors: