Sub cloning

Steven Sullivan sullivan at gwis2.circ.gwu.edu
Fri Aug 1 10:41:11 EST 1997




L.D. Ofner (BMBLDO at leeds.ac.uk) wrote:
: Hi everybody,

: I've been trying to sub clone a full length cDNA sequence (4.5kb) into an 
: expression vector for ages now with no luck and it's driving me mad ! There 
: are EcoR1 sites at either end of the insert in its original vector which I'm 
: using to cut it out (there are no other restriction enzymes that I can use 
: for directional sub cloning which won't cut the insert sequence or the 
: expression vector in the wrong place). I've performed the necessary 
: restriction digest for cutting the insert out and cutting the expression 
: vector, purified both from agarose gel using Wizard clean up and 
: dephosphorylated the expression vector ready for ligation. I've set up 
: numerous ligations with differing ratios of insert:vector (including having 
: insert in vast excess) and have had loads of colonies growing on plates. But 
: when I prep up the DNA ( usually pick at least 10 colonies) and check by 
: restriction digest, none seem to contain insert ! I've checked that the 
: vector is dephosphorylated by tranforming my competent cells with it and I 
: get no resulting colonies on plates. Help ! Where am I going wrong? I'd be 
: grateful for any advice or suggestions (soon before I go totally insane !).

I have a similar problem (see the thread called 'Ligation Mystery'). 
Several answers have been offered:  1) UV light is messing up your DNA
during the gel purification step (see Biotechniques, November 1996, which
suggest adding guanosine or cytosine to the gel as a protectant) --
however, I'm still not clear on how you could get blank control plates if
this were the problem; and surely we've all previously successfully
ligated DNA that has been gel purified;  2) your water is 'bad'; 3) 
dephosphorylated ends tend to 'go blunt', allowing them to religate. 
Myself, I'm going to stop dephosphorylating (it's not really necessary for
my vector) and try the UV protection method anyway and see if it helps.




More information about the Methods mailing list