Recording dynamically molecular interactions with freeze clamp

[herveboston]

guys,

 

I'd like to discuss this idea (on youtube on the link below) to record which molecule(s) is(are) interacting with which ones when two kind of cells met. So far, in any case, at least one of the partner has to be know. The game then consist to try to identify who is binding to it (blotting, screening, labeling). Most of the time, it required firstly purification of known proteins. Here the idea is to

reverse the mapping of cell-cell, cell-protein and protein-protein interaction by freezing, linking them with UV and then sequence the attached parts after isolation to identify the interacting partners. I use freeze clamp to freeze ionic channel so, they can be labeled with photoactive probes.

My hypothesis is to use photoactivatable probes in the media where are the cells. When they become spatially very close to each other (example during an invasion), at their interfaces, some probes will be trapped. If the surfaces are complementary geometrically, the bond will be stronger (?). When the media is instant frozen, each partner will stay in its exact position at a precise moment of the molecular events happening. Nitrogen freezing enables the fast freezing. So, if the arms of the probe is close enough of each surface molecules, the UV irradiation will bind them together covalently. If the trio (part of cell A, part of cell B and the photoactive linker) are labeled with specific "dyes" or labels (radio-elements?), the all process will be to track any band on gel having the three signals. These trio bands having the label of the three markers will be purify and sequenced (liquid or Mass spec). The obtained sequence enables the identification of the partner in data base. Therefore, no need to know the interacting partners before sequencing them...

Thank you for your enlightenment

HB

.

 

 

link youtube: http://www.youtube.c...h?v=5QMrb-d9NaM

[mississippichem]

So, if the arms of the probe is close enough of each surface molecules, the UV irradiation will bind them together covalently.

 

Would you mind explaining this part in a bit more detail? Photodissociation followed by a bond forming event can be some difficult to chemistry to make happen predictably. The nature of the photodissociation will be highly dependent on the reactivity of the molecule and the immediate neighborhood of functionality present. For example, the presence of a nearby phosphorus atom (nearby to your photo-labile group) will make this difficult.

 

Could you perhaps provide an explicit pair of molecules as an example?

 

I didn't watch the video so forgive me if this has been addressed there.

[herveboston]

Hi Mississipichem,

 

The idea here is NOT to look for a functionality but molecules happening to be spatially close to each other during a biological process like a cell invasion by a parasite (I used in the video the example of red cell invaded by a malaria parasite) and then to follow the link during the time. The hypothesis is that participating interactions should have specific pattern during the process while not participating interactions would be random. The linker being added in the media where the process is taking place. After being instantly frozen under very thin film and irradiated with UV, the binding ends of the linker get activated and will bind to surrounding molecules. Reactive molecules spatially close will be covalently attached. The goal then will be to track down which molecules from entity A (let's say a cell) are linked to molecules from entity B (another cell, a parasite, etc...) during this process. Once isolated (sequential enzymatic digestion, PAGE, purification, etc....), the bands carrying the three signals (label of A, label of B and label of the linker) should be the linked ones. The identification being ultimately done by liquid sequencing or mass spec associated to a data base. The challenge of differential reactivity of the linker to specific amino acids is an issue only if you are considering to pick all interacting molecules. I don't think this could hardly happen. Picking what could be pick could be already great. An example of such linker (on the video) is TID used to labeled hydrophobic pockets in studies the gating of ion channel receptors. I'll let you watch the video and we can talk then.

 

Thanks

 

herveboston