[DIYbio] Re: TLR Biosensor

Hi Simon,

Immunobiology grad student here.  Sounds like a very interesting although potentially challenging project!  It's great to explore the possibilities.

Regarding "background" dimerization of the split-GFP:  This is a concern, but careful design and testing could avoid it.  Some TLR's exist as monomers before they see the ligand, some exist as dimers.  I think TLR2 and TLR6 are in the category that exist primarily as monomers, but there's still likely to be some random interaction.  I would try to take advantage of the change in protein conformation that occurs when TLRs bind ligand, which brings the cytoplasmic TIR domains closer together.  The length and "floppiness" of the linker to the GFP-fragment could be optimized so that interaction between split-GFPs is unlikely unless ligand binding occurs. 

A few links:
http://www.ncbi.nlm.nih.gov/pubmed/24419035 --Overview of molecular mechanisms of TLR ligand binding
http://www.ncbi.nlm.nih.gov/pubmed/19378018 ---Book chapter exploring ideas very similar to yours, although they don't propose bacterial expression.  They discuss both split-GFP and FRET methods.  From personal experience, FRET gives you beautiful data when it works but can be a major PITA!

Some speculation...I'm not too familiar with function of animal membrane proteins in bacteria, so I can't comment on how feasible that would be.  However, if you can't get bacterial expression to work, maybe consider plant cells?  Plants have their own set of pattern-recognition receptors (PRRs) for bacteria, virus, etc...not TLRs, but similar LLR structure.  (Toll-like-receptor-like-receptors?)

Final comment...Blenderkid's IEEE link shows a good point, immune cells are already exquisite little contamination sensors.  I routinely treat mouse dentritic cells with 1ng/mL lipopolysacharide, and 6 hours later they have changes in gene expression, surface markers, and even cell shape.  This is just a few thousand molecules of LPS per cell!

Cheers,
Mike


On Tuesday, February 3, 2015 at 1:32:17 PM UTC-5, Simon Rose wrote:

Hi Blenderkid  and Gavin, sorry about the delay in responding,   I've been offline for a few days !

In response to Blenderkid's question, yes I intend to try to couple the TLR's  to a cell. Initially I thought of using e.coli as a chassis,  and as TLR2/6 respond to gram- positive bacteria and e.coli is gram negative, self- detection should not be a problem. It would be something on the lines of a little "machine" for sensing gram-positive pathogens. However, I am very unsure if the transmembrane domains of both TLRs are even capable of being anchored in the cell -wall  of a gram negative bacterium, so I might have to use yeast as a chassis.

The TMs of TLRs 2 and 6 are both 21 AAs in length by the way. There is some information about them on PLOS1. Apparently isolated TMDs are  capable of oligermerisation independently of any PAMPS detected by the extracellular portions (PLOS 1   November 2012 | Volume 7 | Issue 11 | e48875 )

As  for detection one idea would be to extract  the TLR from say, HeLa ( assuming they're not too badly mutated) having tagged the primers with histidine codons, ligate into   an expression vector and then detect any expressed protein with anti-6His antibodies, just to see if anything happens.

As for  the problem  with split GFPs flopping around and dimerising independently, well this should not be a problem if the TLR GFP constructs are anchored in a cell membrane/wall. There has actually been some fairly recent work on fluorescent biosensors here;

  Eiji Nakata, FongFong Liew, Shun Nakano and Takashi Morii (2011). Recent progress in the construction   methodology of fluorescent biosensors based on biomolecules, Biosensors - Emerging Materials and Applications, Prof. Pier Andrea Serra (Ed.), ISBN: 978-953-307-328-6,rey ss b (

These ideas are still in progress by the way.

Its nice to have an international community to help out. The immune system fascinates me, by the way which is why I'm doing this! Thanks for the responses so far!