The resulting plasmid has two nicks (single-strand breaks) at the positions lacking the 5'-phosphate ends. This results in less transformants, but once inside the bacterial cell, the DNA repair machinery will take care of that. Thus, you will get normal plasmid DNA back out of your colonies.

Dear Ruckert: thank you for your answer, but this is not true, we got transformants, however, the ends were not normally ligated. Therefore, I ask what will happen when blunt ends are ligatgated by T4 ligase, it seems T4 ligase doing something at blunt ends before ligation.

No, it usually does not. I have done dozens of ligations of that type (with and without kits) and never observed any discrepancy.

What kind of modification did you observe that you attribute to the ligase?

The only activity attributed to T4 DNA ligase is, as stated by Christian, catalysis of phosphodiester bond formation between closely apposed 5' phosphate and 3' hydroxyl ends of duplex nucleic acids. I have done a variety of blunt end ligations also and observed the expected products, albeit at lower frequency then would be seen for sticky end ligations. Maybe you could provide more information about the fragments you are ligating and products you are getting- that could help with troubleshooting.

Because there is no 5' phosphate between two blunt DNA ends, it seems T4 ligase manage to produce a 5' phosphate before ligation. Perhaps not the nicks (single-strand breaks) is ligated in E.coli, but the 5' phosphate os produced in vitro and ligated before transformed into E.coli. Therefore, I need to know what will happen in this situation?

Dear Lianqwei.

As Christian said you will get a double nicked plasmid that will be repaired by the bacteria.

The phosphate used for that ligation events are the ones provided by the plasmid that indeed has two 5`-phosphates after a restriction with blunt cutting enzyme like EcoRV.

Thank you everyone, what you said is commonly thought so. However, the vector we used is produced by PCR, not cut by restriction enzyme, therefore, it has no 5`-phosphates. What will happen?

That raises the question why you did this (and did not phosphorylate the insert).

Anyway, under these circumstances, the T4 ligase will do nothing at all. As you got some transformants, the most likely explanation is that you transformed two pieces of linear DNA (vector and insert) and that these were recombined in the bacteria by the DNA repair systems. That will most definitely alter the molecules at the recombination sites.

Thank you Ruckert, can you provide some data to prove "the T4 ligase will do nothing"? because after T4 ligation, the number of transformants rises. It seems that T4 do something.

Without a 5'-phosphate, no ligation can take place. That is common textbook knowledge in which the proper references can be found, the most common of which might be http://www.ncbi.nlm.nih.gov/pubmed/4377758.

One possible explanation for your observation: The T4 ligase can still attach itself to the blunt end (see, e.g,. http://www.ncbi.nlm.nih.gov/pubmed/?term=9153309). That might protect the DNA against degradation by exonucleases. You might want to try the N-∆80 described in the aforementoined paper if you want to test this hypothesis.

Did you amplify your plasmid by PCR? instead of growing bacteria and maxipreping?

Why?