Photoswitchable conformational change, azobenzene, unatural amino acids, etc.

[Bryan Bishop]

1) Unnatural amino acids have been constructed that incorporate azobenzene. These amino acids seem to be compatible with ribosomes and they can be incorporated into the output proteins.

Photoswitchable click amino acids: light control of conformation and bioactivity.

http://www.ncbi.nlm.nih.gov/pubmed/21998087

Genetically encoding photoswitchable click amino acids in Escherichia coli and mammalian cells

http://www.ncbi.nlm.nih.gov/pmc/articles/PMC4051619/

In situ formation of an azo bridge on proteins controllable by visible light

http://pubs.acs.org/doi/abs/10.1021/jacs.5b06234?journalCode=jacsat

Photosensitive GFP mutants containing an azobenzene unnatural amino acid

http://www.sciencedirect.com/science/article/pii/S0960894X14013432

2) Also azobenzene (and friends) can be incorporated into ribozymes and DNAzymes. 

Light-driven DNA nanomachine with a photoresponsive molecular engine

http://pubs.acs.org/doi/abs/10.1021/ar400308f

Photoswitch nucleic acid catalytic activity by regulating topological structure with a universal supraphotoswitch

http://pubs.acs.org/doi/abs/10.1021/sb300120n

http://gnusha.org/logs/2016-07-15.log

http://gnusha.org/logs/2016-07-16.log

- Bryan
http://heybryan.org/
1 512 203 0507

[Kent Kemmish]

Is this the way?
1. design/construct an mRNA or cDNA display library of 1e12-1e14 mutant polymerases (diversity should approach 1e12-1e14, with 1e3-1e4 copies of each mutant present) with photoswitchable amino acids (important thing is that each mutant is covalently attached to its code, whether in mRNA or cDNA form, or whatever, best is to begin with something single-stranded)
2. first round amplify with intense white light
3. size select double-stranded constructs (everything that worked)
4. second round in darkness
5. counter-select ds constructs
repeat 2-5 for 20-30 rounds
voila, you'll have mutant polymerase that only work with white light.
extend basic concept to four or five colors

This idea framework has been around for at least several years (independently thought up by at least several people including me...) so it begs the question, why not? Have stealth efforts gone down that road and failed?

"It takes a great general to see the obvious"

I fully realize the above "simple recipe" for getting a light-programmable base-specific polymerase (and you could obviously do exactly the same thing with terminal transferase mutants and just size select anything that gets longer rather than double stranded... the variants write themselves... in more ways than one) could be a year of painful sweat at the bench, but at first approximation (hell, at least eight years ago!) it's what I would do with funding/freedom/focus to try to solve synthesis rather than write emails about (which is all I do other than developing possibly relevant nanopore tech--but I'm not sure nanopores are needed to solve this, unless you need to sequence between rounds...)

[Bryan Bishop]

On Sat, Sep 17, 2016 at 1:28 PM, Kent Kemmish <kemmi...@gmail.com> wrote:

This idea framework has been around for at least several years (independently thought up by at least several people including me...) so it begs the question, why not?  Have stealth efforts gone down that road and failed?

Well, some of us just take a while. There are some initiatives from my end in this direction. One is a draft paper that gives a review of many of these methods. Second is that there is some funding available for this apparently, although it's still early stage. I think a publication (review) would be a good first step. I think the problem was that 8 years ago I didn't have as much of my own funding to play around with :-).