Loss of protein expression by transformed bacteria - (May/17/2015 )

I am in desperate need of advice.

I did transformation of BL21(DE3)pLySs, Rosetta 2, Rosetta gami and T7 Express LysSIq cells last year with the my genes of interest cloned in pET20+b vector. After almost 4 months of trial and error, I finally got the positive colonies which I confirmed with WB. This year I started to work on purification of my protein. I took small amount from my glycerol stocks and streaked them on fresh LB plates with 100ug/mL of carbenicillin and picked 5-6 colonies each time. Did pre-culture and then started the process of large scale culture and purification. I didn't get any protein. Then I started to check the expression in 3 ml cultures from colonies picked from plates and now I have almost checked all my bacteria and none of them show expression. I used IPTG from 0.4mM to 1mM and did different time and temp conditions. Any advice? 

Can someone help with loss of expression in transformed bacteria?. 

Could you provide some more information about your protein? The 4 E.coli strains you are using for transformation are generally used to improve expression efficiency in difficult to express proteins. Does your protein have unusual codons for an E.coli system? Is it known to be toxic?

Also, it is usual to re-transform these strains rather than to use older glycerol stocks. I don't know the mechanics of the problem, but experience has shown that re-transformed cells perform better than glycerol stocks. I'd suggest re-transforming, pickiing single colonies, and moving forward.

Agreed with phage...always do a fresh transformation for every expression. Make a big stock of competent BL21 cells, aliquot, and freeze. Transform expression plasmid for each expression and you'll always have robust and reproducible results.

Also, no need for clonal cultures when doing expression imo. Scrape a whole bunch of colonies from the freshly transformed plate of cells and use that to inoculate your cultures.

Selecting a single colony is the "traditional" way, and many PI's will be adamant that you need clonal cultures for expression when in reality that's not ideal because different colonies will express at slightly different levels (I've tested this), and if you are unlucky and happen select that one bad colony, then you're putting yourself at a disadvantage. And testing multiple colonies to find the best each time you do a transformation/expression is a waste of time.

By scraping many colonies you get the population average expression every time. Hence near perfect reproducibility. Doing it this way, my yields are always within 5-10% for each prep, and they never don't work unless there is user error. I used to be a "clonal expression" guy, but I am now converted and will never go back.

Yes. You should  re-transforming, picking single colonies, and then move forward.