Thanks very much again and would like to further your suggestions here,
Ah. Well, I'd suggest that the major problem might be the size of the vector, rather than the ccdB insert.
I thought potential size issue as well, I tested another ccdb vector (19kb) with inserts that replaced ccdb sites as a size control vector. Similarly, this 19kb vector with ccdb cannot transform DB3.1, but vector/insert without ccdb had very high tranformation, indicating high competence of the self-made DB3.1 cells.
It would be worth checking to see if the competence of your cells was high without the ccdB insert. I suspect that you will see similar transformation difficulties with such a large vector. Are you using electroporation or chemical transformation?
I made electro-DB3.1 competent cells.
If you prepare a large vector without ccdB, then you could also check the transformation efficiency of other strains, some of which are optimized for large vectors. Have you thought of replacing the ccdB selection with a visible phenotype, such as LacZ (on S-Gal plates) or GFP/RFP proteins?
I would definitely need this ccdb sites for LR reaction, that is why my insertion modification is designed in places other than ccdb sites.
I appreciate your further insights based on the above discussion....