Designing XOR Gates - two campus approach

Davidson XOR Biological Design

List of auto-inducers and their catalog numbers.

Davidson Approach

Here is an idea Malcolm and Laurie developed.

Everyone please look at this and ask questions and find holes in it now so we don't waste time building something that won't work.

The idea is to have two mirrored halves of the system. LasR is regulated by PAI-1 {3-oxododecanoyl-HSL (3OC12HSL)} and LuxR is activated by AI-1 {3-oxohexanoyl-homoserine lactone (3OC6HSL)}. There is a potential problem in that the Lux half is more likely to get positive feedback than the Las half. This MAY not be a problem because 0/0 is leaky so we put a weak RBS to minimize leaky protein production. Also, if we add AI-2 and AI-1 is produced by leak, then the entire system shuts down. The repressor site is located between -35 and -10 of the promoter. The activator binding site is upstream of -35. This has been documented by Egland and Greenberg

Oligos_to_Build: Sequences we will need to make this XOR gate.

LIST OF Km Values

• Km of 3OC12 for LasR is 1 nM (Egland and Greenberg, 2000)
  
• Km of 3OC6 for LuxR is 100 nM (Urbanowski et al., 2004)
  
• Km of LasR* for operator/promoter is 11 pM (Schuster et al, 2004)
  
• Km of LuxR* for operator/promoter is 10 nM (Kaplan and Greenberg, 1987)
  
• Km of cI dimer for OR1 and OR2 is 10 pM (King et al., 1993)
  
• Km of Mnt tetramer for binding to half operator/promoter is 50 nM and binding whole operator is 1 nM (Berggrun and Sauer, 2001)
  
• Km of Lsr for its binding site is X nM (no data available)
  
• Km of AI-2 for LsrR is X nM (no data available)
  
• Km of IPTG for LacI is 1.3 µM (Gilbert and Muller-Hill, 1966)
  
• Km of LacI for its binding site is 10 pM (Gilbert and Muller-Hill, 1967)
  
• Km of LacI-I12 for its binding site is 0.13 pM (calculated from Schmitz and Galas, 1980)
  
• Km of LacI-X86 for its binding site is 0.13 pM (calculated from Schmitz and Galas, 1980)
  
• Km of LacI-I12_X86 for its binding site is 0.001 pM (calculated from Schmitz and Galas, 1980)
  
• Half life of molecules in E. coli: 2 minutes for mRNA; 1 molecule is 1 nM concentration (Fang Ph.D. Thesis, 2007); translation rate 15 amino acids per second and transcription is 40 nt per second (from Genes VII by Lewin).

Missouri Western XOR Biological Design 1

These two XOR circuits are designed to complement each other. Each recieves a cell-to-cell signal (AI-1 or AI-2) and a chemical signal (IPTG or AHL) and processes it into a cell-to-cell signal. Colonies that output AI-1 would alternate with colonies that produce AI-2. The input message to be hashed could be encoded by the presence or absence of the chemical signals, which would also alternate.

Design Variables

1. strength on RBS for each of the coding sequences (eg. RBS for enzymes could be weak)

2. order of coding sequences (eg. enzymes could be second for lower expression level)

3. identity of repressors (Davidson is building three different LacI repressors)

4. type of degradation tag on proteins

Things to Do

1. make a list of the parts needed and their building status

2. design, order, and clone 4 hybrid promoters

3. test the 4 hybrid promoters with GFP or RFP outputs

DNA Sequences of Relevant Parts

Parts from the registry

Oligo design for XOR Hybrid Promoters

Parts needed for XOR

Build XOR Gate

Test XOR Hybrid Promoters

Above - Input of AI-1 or IPTG turns on production of AI-2 by LuxS. Input by both AI-1 and IPTG allows production of the repressors cI and Mnt, which repress both transcription units. LuxR and LacI are constitutively expressed.

Above - Input of AI-2 or aTc turns on production of AI-1 by LuxI. Input by both AI-2 and aTc allows production of the repressors cI and Mnt, which repress both transcription units. LsrK, LsrR and TetR are constitutively expressed.

Combined Davidson/Missouri Western XOR Biological Design

A strength of the Davidson XOR design above is that each of the two signalling molecules used is foreign to E. coli. A strength of the Missouri Western design above is that is uses two complementary XOR gates. The design below combines the two approaches. Each XOR circuit receives a cell-to-cell signal (AI-1 or P-AI-1) and a chemical signal (IPTG or AHL) and processes it into a cell-to-cell signal. Colonies that output AI-1 would alternate with colonies that produce P-AI-1. The input message to be hashed could be encoded by the presence or absence of the chemical signals, which would also alternate.

Input of AI-1 or IPTG turns on production of AI-2 by LuxS. Input by both AI-1 and IPTG allows production of the repressors cI and Mnt, which repress both transcription units. LuxR and LacI are constitutively expressed.

Input of AI-2 or aTc turns on production of AI-1 by LuxI. Input by both AI-2 and aTc allows production of the repressors cI and Mnt, which repress both transcription units. LsrK, LsrR and TetR are constitutively expressed.

Missouri Western XOR Biological Design 2

XOR Based on Tryptophan Anabolism and the TrpR Repressor

Notes: LacI could be the new LacI X86+I12. Also, the output gene should be LuxI, not LuxS. With LuxS, the second cell has to have the Las system in place of the Lux system (JB/AMC).

The idea here is that there are two different XOR gate clones. One takes input of AI1 and IPTG and outputs AI2. The other takes inputs of AI2 and IPTG and outputs AI2. These two clones could be alternated in a pathway of colonies (TE/AG).

Missouri Western XOR Biological Design 3

This circuit connects the two transcription units through synthesis of AI-1 and LuxR, which combine to repress both units. It responds to AI-2 and outputs AI-2. An analogous circuit could have input of AI-1 and IPTG and output of AI-1. However, these two could not communicate with each other.

Above - Input of AI-2 or IPTG turns on production of AI-2 by LuxS. Input by both AI-2 and IPTG allows production of LuxI and LuxR, which combine to repress both transcription units. LsrK, LsrR, and LacI are constitutively expressed.

Davidson Ampicillin Communication: time delayed cell growth

The very first colony has a high copy number plasmid that is Amp and Kan resistant (^R). As this colony grows over time, it will digest ampicillin in an increasingly larger circle shown by radiating circles of blue. The subsequent colonies have a lower copy number plasmid and have a promoter-less version of Amp^R in addition to the XOR construct. The Amp^R coding will have either its native RBS or one we give it.

Davidson Growth Layouts

To enhance the unidirectional flow of Amp^R, we could either grow the cells on a slant or create a vertical stack of agar plugs. The thickness of the plug would be determined by the thickness of the plates we pour. This may or may not help with the diffusion of Amp^R but it is easier to do than microfluidics.

---