Team:Edinburgh/Notebook/Project summary/01 July 2008



02 July 2008

E. coli cell lysis

Genes from bacteriophages can cause bacterial cell lysis.

Simple lytic phages, such as the single-stranded (ss) DNA phage φX174 and the ssRNA phages, have a single lysis gene (prototypical genes being φX174 E, MS2 L and Qβ A2). φX174 E seems to be the best characterised at the moment, and encodes a single lysis protein that inhibits a specific step in murein biosynthesis. Fewer than 500 molecules of E are present at the time of lysis. Lysis by E requires continued host cell division. (We can't use the system that Paris used to make their "germline" non-divisible.) The DNA sequence for E is:


translates to:


01 July 2008

Lysis: An alternative to cellulase secretion page updated.

Paris '07designed what they call "synthetic multicellular bacteria" (SMB). This involved reproductively capable "germline" cells which were dependent on diaminopimelate (DAP), a metabolite produced by reproductively incapable "somatic" cells. Low concentrations of DAP triggered the transcription of Cre-recombinase (under the influence of DAP-sensitivepromoter), which caused the removal of a gene required for bacterial reproduction by site-specific recombination (SSR), and resulted in the activation of a gene for the production of DAP by them.

We could adapt this system to cause some cells to produce cellulases and then lyse to release them into the medium. However, a simpler method could by-pass the need for SSR and could be as follows:

Our cellulase genes could be under the influence of a promoter sensitive to glucose concentration (activated by low [glucose], inhibited by high [glucose]). Low glucose would, therefore, cause the production of cellulases in some cells. We could engineer a reporter gene just downstream of a cellulase gene (under the influence of the same promoter), which could act as a transcriptional activator for lysis genes. So low glucose concentration would cause some cells to produce cellulases and the lysis gene transcription factor. Thus the cells would lyse, releasing the cellulases into the cellulose growth medium. The cellulases would work ex vivo to break down cellulose into glucose, which could then be taken up by other cells and converted into starch until the glucose is depleted to insufficient levels. This seems to be one of those ideas that, if it works, would be very "elegant", an abstract term that biologists seem to love! It might also provide some more interesting modelling opportunities than the very linear metabolic pathways, and we shouldn't need so much data to model this idea. (I still think that secretion of cellulases would be better though - it would be more straightforward, if possible.)

Next step along this train of thought: Find a promoter that is repressed by glucose!

Glucose-sensitive cellulase synthesis

E. coli has loads of glucose-senstive genes, most of which are positively regulated by cAMP:

The action of adenylate cyclase is repressed by glucose. Low [glucose] results in an increase in the activity of adenylate cyclase, which converts AMP into cyclic AMP (cAMP). cAMP associates with the cAMP receptor protein (CRP, also known as the cAMP activator protein, CAP). CRP¿cAMP can then associate with the CRP binding site upstream of the target gene promoter. CRP¿cAMP then interacts with RNA polymerase, resulting in transcription of the gene.

Some examples of genes controlled in such a way are the rpoH(¿32, the heat-shock ¿-factor), the sdhCDAB operon and the genes of the CAP regulon. The best characterised system seems to be the lac operon (part of the CAP regulon).

The lacoperon has its CRP binding site at -70 -> -55, where CRP¿cAMP binds as a dimer and interacts with the C-terminal domain of the ¿-subunit of the RNA Polymerase. We could take this sequence and add it to the same position upstream of the promoter of an artificial cellulase operon.

There are a couple of other things to think about though:

The lac operon has its activity repressed by the LacIrepressor, which binds to the operator, preventing the polymerase from binding to the promoter. Allolactose (an analogue of lactose) displacesLacI. - What this shows is that we do not want to copy the operator! Other catabolite-sensitive promoters bind CRP¿cAMP in two locations, as two dimers. This could potentially be better for us. Needing 2 dimers to bind would make initiation of the transcription less sensitive. Searching EcoCycfor CRP¿cAMP transcriptional dual regulator regulated genes gives some 200 results. I propose that we want a gene whose activity is only controlled by CRP¿cAMP, of which there appear to be five. The best candidate out of these seems to be cstA, the gene for a peptide transporter induced by carbon starvation. The CRP¿cAMP transcriptional dual regulator is located at -89.5 from the transcription start site.

This gives the sequences:

CRP-binding site: act cggttaaCGG AGTGATCGAG TTAACATTGt taagttaaa = 628968->629009 Promoter: tca actccgattt acatggttgc tgtgttgtta aattgtacaa agatgttata gaaacaaAat gtaacatctc tatggaca = 629018->629098

Put into context: CAGGAAAact cggttaaCGG AGTGATCGAG TTAACATTGt taagttaaaT ATTGGTTtca actccgattt acatggttgc tgtgttgtta aattgtacaa agatgttata gaaacaaAat gtaacatctc tatggacaCG CACACGGATA ACAACTatgA ACAAATCAGG GAAATACCTC (628970-629140)

Where the "A" represents the start of transcription and "atg" is the translation start codon.

We could add this sequence (up to the start codon) to our cellulase operon.

References Snyder, L. & Champness, W., 2007, Molecular Genetics of Bacteria (3rd Edition) (Book)