Protocol Online logo
Top : Forum Archives: : Molecular Biology

T4 Ligase problem - Please help me, (Nov/11/2008 )

Pages: 1 2 Next

Hallo everybody,

My name is Nicola, I'm Italian and I work in a molecular pharmacology lab.

I'm trying a ligation reaction since one month and I still don't have my plasmid....after the transformation I always recover empty dishes, no colonies, it's kind of frustrating, even because this reaction is supposed to be quite easy and I'm using tons of precautions!

Theorically, my ligation is:
"VECTOR" ---- pcDNA 3.1 (-) double cut with KpnI and XhoI ---- 5400pb
"INSERT" ---- Sequence YYYY extracted from another plasmid, cut with KpnI and XhoI ---- 2000pb

As you can see, both the fragments I want to ligate are sticky ended and the extremities are completely complementary.

The protocol I used is the following:

1- Restriction enzymes reactions on the two plasmids
2- Agarose gel separation
3- Extraction from gel with columns and kit buffers
4- Quantification of the extracted bands
5- Ligation with T4-invitrogen normal (not fast) ligase kit

The ligase conditions were:
- 15 fmol plasmid
- 45 fmol insert
- 4ul buffer 5X
- 1unit T4-ligase
- water nuclease free to 20ul
- 16 degrees Celsius overnight

Then transformation and.....nothing!!!

Any idea? Can you please suggest me another protocol or another technique?

Please help me, I desperately need the new plasmid for my experiments

Thank you in advance



You shall check every involved details of the procedure and stuff itself in the first place. U know, if one thing went wrong in the process, your whole experiment fell apart. sleep.gif



May i ask if you have done any ligation before? The way to troubleshoot is different.

Your protocol seems normal and no problem. Below are a few suggestions:

1. Check transformation efficacy by using a control. Otherwise please confirm the competent cell efficacy by doing any type of plasmid propagation.

2. Increase your vector to insert ratio. 1:4, 1:5... Even though i do not think this is an all-or-none step.

3. What else...hmm...try to confirm your antibiotic selection tongue.gif . Check antibiotic resistence gene on plasmid and use same antibiotic on the plate.

Wish you good luck smile.gif


Sorry, but the conversion from fmol to ng vector and insert is too much right now. Ligate 50ng vector in 10 ul with the required amount of insert (~40 ng, I think).

In order to test your ligase, try addingsome to DNA ladder (in ligase buffer, of course). Treat at RT for 15-20 minutes, then run on a gel. If you don't have a change in the ladder (everything sfited up), your ligase is a (the?) problem.


First of all, thank to everybody.

So, I tried many ratio, but I'm going to increase the insert ratio.

I didn't make the control trasformation with native pcDNA3.1 because I always used the same transformation protocol with the same DH5alpha heat shock competent E.Coli line...and it always worked! So I don't know if I should do also the control transformation.

The resistance of the pcDNA3.1 is AMPICILLIN, I made this mistake once 2 years ago and my Professor already made fun of me for this rolleyes.gif !!! I felt like an idiot when it happened dry.gif !!!

I didn't check the activity of the T4 ligase. We have the Fermentas HindIII Lambda. That's an interesting control...I'm going to try it.

A friend here thinks that the problems could be the extremities of the DNA fragments. Do you think the fragments extremities can suffer the extraction with commercial kits giving some problems to the ligation reaction?

Thank you in advance and all the best!



Do you dephosphorylate your vector? You don't mention this step, but it's crucial.

-Ginger Spice-

F*=#, no, I didn't dephosphorilate it! I asked about this to my professor....and he f*=###$g told me it wasn't necessary!

Do you think I should do it?

Thank you very much for your advice.


QUOTE (Ginger Spice @ Nov 12 2008, 11:08 AM)
Do you dephosphorylate your vector? Yo don't mention this step, but it's crucial.


QUOTE (NikiITALY @ Nov 12 2008, 11:21 AM)
F*=#, no, I didn't dephosphorilate it! I asked about this to my professor....and he f*=###$g told me it wasn't necessary!

Do you think I should do it?

Thank you very much for your advice.


I ALWAYS do, even when cloning with two different restriction enzymes. Imagine that digestion of your vector isn't quite complete: some have been linearized with only KpnI or only XhoI, and these can re-circularize. You can use gel purification to separate linearized vector from uncut vector, but you don't know for sure that each vector molecule has been cut once with each enzyme. This doesn't explain why you have no colonies, but I think it's still an important step.

Regarding the DNA ends, how close together are the cut sites in your vector? If they are quite close, your efficiency of double-cutting will be very low, and hence your efficiency of ligation will also be low.

To check how good your ligase is, digest a vector with a single enzyme and gel purify it. Incubate some linearized DNA with ligase and some without, then transform some of each into E. coli. If your ligase is good, your cut-ligated DNA will give lots of colonies and your cut DNA should give none.

You can also check each restriction enzyme by performing a single digest of plasmid DNA to see if it is efficiently linearized by your enzymes.

And just once, do a ligation control to make sure your technique is good.

I hope you have colonies soon!


-Ginger Spice-

Ginger, in effect, in my pcDNA3.1 the digestion sites are very close!


QUOTE (NikiITALY @ Nov 12 2008, 07:58 AM)
Ginger, in effect, in my pcDNA3.1 the digestion sites are very close!

Normally, the efficiecy of the Restriction sites from MCS close to each other will be fine. The commercial vector have already been designed properly.
However, to confirm its efficiecy, you may connect the link below:

Best wishes

-Jeff Yuan-

Pages: 1 2 Next