More on m13 and say HIV

OK some background. M-13 is a virus which infects e. coli. It has a number of very useful features one of which is that it displays protein sequences inserted into one of its coat proteins. Another is that, since it does not kill the e-coli it infects, a huge amount of M-13 can be produced by an infected e-coli. The key for the main use of M-13 (irrelevant here) its genome is single stranded.

There is a (I forget the sites) an established technique of using m13 expression to find proteins that stick to something. In particular, the very first application was to find human immunoglobulin heavy chain variable regions which stick to a target antigen.

I have some thoughts on a non standard way to do this (I won’t describe the standard way).

plan. Make a M-13 which presents the target antigen and has a heat sensitive (or nonsense) mutation in an essential gene. Make a library of M-13 which express human heavy chain variable regions with a heat sensitive (or nonsense) mutation in another essential gene. Mix, dilute and infect e. coli at high temperature (or non nonsense suppressing). Repeat always adding the target antigen M-13.

Now a little more. to find a heavy chain variable region which binds to the antigen and knocks of another binding protein call it, say CD-4. Here first express target antigen as above. Express CD-4 on an M-13 with a dominant lethal (say streptomycin sensitivity or a heat sensitive coat protein).

First mix the target antigen M-13 and excess dominant lethal CD-4 M-13 and check that successful infection of e. coli is very rare (hah check indeed work for years more likely). Then repeat adding the mix of heavy chain variable region expressing M-13.

If the dominant lethal CD-4 M-13 coinfecting with target antigen M-13 bit works then it could be used to screen for monoclonals which compete off the target antigen.