Biological Turing Machine Idea

Biological Turing Machine Idea Kurt Whittemore Here is my idea for implementing a Turing machine in a biological cell. A Turing machine is a theoretical computation machine invented by Alan Turing in 1937 [1]. These theoretical machines serve as the foundation for the modern computers that we use today. Implementing computational systems in biology may give humans the ability to program biological cells to store and manipulate complex information in order to accomplish different tasks. Cells already do this, but not always in a way that is understandable by humans. Several different research groups have already tried to implement computational capabilities in biological systems. However, most of these efforts have resulted in biological systems that can tackle one very specific problem, and these systems are not capable of performing general computation. In other words, these systems are not Turing complete. For example, one group has used DNA to find solutions to the Hamiltonian Path Problem [2], while another group has used DNA to find solutions to the Burnt Pancake Problem [3]. Many different teams in the iGEM (international Genetically Engineered Machines) competition have created genetic networks that can act as switches, inverters, etc. [4]. Most of these systems require the constant production and degradation of protein, which consumes energy. Another research group has used invertases to flip segments of DNA essentially turning them “on” or “off” [5]. This system does not require the constant input of energy. All of these systems are quite clever, but from what I understand, none of them are Turing complete. The system I propose here should be Turing complete. However, I'm not sure if it could actually be built. This system depends on the ability to splice pre-mRNA transcripts into desired mRNA transcripts. The system also requires the ability to switch different splicing machinery on and off. Zinc finger transcription factors are another critical component of this idea. Zinc finger proteins are proteins that consist of modular “fingers” which can bind to different segments of DNA [6]. Alan Turing's original Universal Turing Machine required several “colors” and “states.” However, a fairly recent competition sponsored by Stephen Wolfram to find simpler Universal Turing Machines has produced a Turing machine with just 2 colors and 3 states [7]. My basic idea here is just to map each feature of this conceptual Turing machine to a biological component or system. A Turing machine consists of a piece of tape containing many different cells. Each cell can be one of two colors. A head in one of three states reads a cell. Then depending on the current state of the head and the color in the cell just read, the head will change the color of the current cell, move to the next or previous cell, and change its state in a certain way. For a 2 color 3 state Turing machine there are just 6 rules required for universality. What are all of the different components in my biological system? The tape is a long sequence of DNA containing many different cells. Each cell is a segment of DNA with a certain address. The address is specified by the sequence in the promoter of that segment that only a certain ZFP will recognize. At the very end of this cell (DNA segment) there is a sequence that can be methylated or not methylated by a ZFP methylase. This sequence is only transcribed into RNA if it is not methylated. This methylation or demethylation corresponds to the color of the cell. The head is representative of the splicing machinery that can be in different states depending on which splicing machinery is activated. This splicing machinery will cut the pre-mRNA transcript from the cell in a certain way to make the system behave just like the conceptual Turing machine. Each pre-mRNA transcript will code for the ZFP for the next address, the ZFP for the previous address, the ZFP for this address, a methylase region for the ZFP, a demethylase region for the ZFP, code to activate splicing machinery 1 (state 1), code to activate splicing machinery 2, code to activate splicing machinery 3, and the methylase region. However, only some of this stuff will survive the splicing process depending on which splicing machinery is activated. Um, yeah, hopefully the diagrams make a little more sense. Here is the diagram for the conceptual Turing machine.

Note that the Turing machine is programmed by changing the initial pattern of black and white squares at the beginning. How can we implement this system in a biological cell? Cell: Segment of DNA in a biological cell Corresponding to a Turing Machine Cell.

The "colors" would correspond to whether the methylase region in the Cell was methylated or not. If the methylase region was not methylated it would have been transcribed. Methylated = Black Demethylated = white

The "state" would correspond to which splicing machinery was activated. The system would behave as follows:

So how would you design the splicing machinery to make a biological system behave like this? I have no idea right now. References 1. Petzold C. The Annotated Turing. 2008. 2. Adleman L. Molecular Computation of Solutions to Combinatorial Problems. Science (1994) 3. Haynes et al. Engineering bacteria to solve the Burnt Pancake Problem. J Biol Eng. 2008; 2: 8. doi: 10.1186/1754-1611-2-8. 4. http://parts2.mit.edu/wiki/index.php/Main_Page

5. Ham TS, Lee SK, Keasling JD, Arkin AP 2008 Design and Construction of a Double Inversion Recombination Switch for Heritable Sequential Genetic Memory. PLoS ONE 3(7): e2815 doi:10.1371/journal.pone.0002815 6. Beerli RR, Barbas CF. Engineering Polydactyl zinc-finger transcription factors. Nature Biotechnology 20: 135-141 (2002). 7. http://mathworld.wolfram.com/TuringMachine.html