Skip to content

10 ways to improve blunt-end ligations

Blunt-end cloning involves the ligation of DNA fragments – usually between a plasmid vector and an insert – whose terminal ends are not “sticky”.

Performing these ligations is notoriously difficult, particularly with large DNA fragments. But it is possible. And in this article I’ll give you some tips that I hope will increase your chances of success.

This type of ligation can be up to 100 times slower (i.e. less efficient) than their sticky-end counterparts, resulting in fewer colonies, and with nothing to force directional insertion of the insert, a proportion of colonies will always incorporate the insert in the wrong orientation.

Despite the problems, blunt-end ligations are worth the effort because – since they don’t require compatible restriction sites – they are extremely versatile so can get you out of a lot of sticky situations in your cloning (pun intended).

Some tips for taming blunt-end ligations

Blunt end cloning reactions are less efficient (compared to sticky end ligations) because of the absence of the hydrogen bonding from sticky ends. This absence means that the interactions between the vector and insert ends is fleeting (look here for an explanation on how ligation works) so, a successful ligation relies on a transient association between 5’ phosphate and 3’hydroxyl groups, caught at exactly the right time by the ligase.

Tip 1: Increase concentrations of insert and ligase

So tip 1 is to increase the concentrations of the insert and ligase as this This effectively increases the likelihood of collisions between the vector and insert ends.

Tip 2: Perform the reaction in two steps

Tip 1 works up to a point. But if you increase the insert and ligase too far you then start to generate constructs with concatamers (multiple inserts). This is because the ligation reaction occurs in two steps:

1. Ligation of one vector end with one insert end (which, as noted above, is relatively unlikely to take place – so the likelihood is increase by higher insert and vector concs)

2. Intramolecular association of the other end of the vector and insert (which is very likely to take place)

If the insert and ligation concentrations are so high that the first reaction is very favored, then it is likely happen over and over within the same molecule – producing concatamers. So the second tip is to beware of raising the insert and vector concentrations too high!

Another way around this is to perform a short 1hr incubation with a high concentrations of insert (favoring reaction 1), then dilute the reaction 1:20 with ligase buffer and continue the reaction for several more hours to allow reaction 2 to occur (thanks to good old Sambrook et al for this one!)

Tip 3: Use longer incubation times

Allowing the ligation reaction to occur over a longer period – up to 24 hours – again increases the probability of two blunt ends bumping into each other and being joined by the ligase. So can make a less favoured reaction more likely to happen

Tip 4: Take care of how you produce the blunt ends

If your plasmid/vector is prepared with a restriction enzyme that creates sticky ends, it can be blunted using T4 DNA polymerase or the Klenow fragment. But since this introduces an extra step – and therefore extra uncertainty, this should only be used as a last resort. A better approach is to use an enzyme that produces a blunt end directly (e.g.  Sma I, Dpn I, Pvu II, Eco RV) then if you verify the digest, you know that you have a blunt end.

On a related note, PCR-generated DNA fragments are always blunt ended, and may be used directly in blunt-end ligations – unless  you use Taq polymerase. Taq adds an extra adenine to the 3’ end of the PCR product, so you’ll need to at a bit of 3′-5′ exonuclease activity (e.g. from a spot of Pfu) to blunt the ends. Again this adds another step and more uncertainty, so is not ideal.

Tip 5: Dephosphorylate the vector

If both ends of the fragment to be ligated into a vector are blunt-ended, then the vector needs to be dephosphorylated to minimise self-ligation. A number of enzymes are available for this step, including shrimp alkaline phosphatase (SAP), calf intestinal phosphatase (CIP) or bovine alkaline phosphatase (BAP)

Tip 6: … and phosphorylate the insert

If the vector needed to be dephosphorylated, as ligation requires the presence of a 5′-phosphate, the insert must be phosphorylated. This can be done by treatment with T4 polynucleotide kinase.

Tip 7: Use a PEG-rich ligation mixture

The theory goes that polyethylene glycol (PEG) “increases” the concentration in the sample by occupying space in the solution. This means the reagents have less space to move about, increasing the likelihood that blunt ends will meet. (Although not without checking this article for some caveats!)

Tip 8: Not working? Make sure it is the ligation that is the problem!

If your blunt ended cloning refuses to work, it is all to easy to jump to the conclusion that it is the “difficult”  ligation that is the problem. But of course, as with any ligation, you need to have a comprehensive set of controls to let you understand where things are going wrong. This article provides an overview of the controls you need.

Tip 9: Persevere!.. to a point

If it doesn’t work first time, try and try again. But if it does not work a third time – and the controls tell you it is the ligation that is the problem – then you make need to reconsider your strategy.

Tip 10: Go around the problem

If you can’t resolve your blunt-end ligation problem – and you have no other way to get the construct you want – a bit of re-engineering may be in order. For example, if you’re going blunt because your plasmid’s multiple cloning site doesn’t have a suitable sticky-ended site for your insert – you can easily add in a new site using something like quikchange. There is always another way!

Good luck with your blunt end ligations — let us know how they go!

Image Credit: woodleywonderworks


  1. Nikolai Scherbak on May 4, 2016 at 12:38 pm

    Well, I would like to make a general comment on Yevgeny’s comment. I think this article on ligation is still up-to-date and going to remain such for a while. There are lots of kits out on the market that simplify your life, if you can spend more money (if you have money) on those kits. However when having money it is cheaper (comparing to kit cost+time/effort) in many cases to order ready cloned constructs from GeneScript or similar. I have done that. But in some cases, for instance when building up a new vector, shuffle cassetes between vectors etc all these kits would not be helpfull that much.
    Thank you for the article and thank you Yevgeny for the summary of alternatives to ligation and advice with agarose gel!

    Örebro University, Sweden

  2. Mohamed on October 12, 2015 at 12:11 pm

    Our problem with this vector that when we make digestion after ligation , we got small un expected fragment rather than our target and by sequencing we may find part of vector between two blunt ends or only primers present or part of our targeted fragment not whole length
    could you help me to solve this problem
    i.e my fragment size between 4 kb and 5 kb

  3. Yevgeny Berdichevsky on January 27, 2014 at 9:16 am

    Dear Alex,
    My name is Yevgeny and I am “ardent fan” of Bitesize Bio and I like its articles and webinars very much. I am sorry, but I was not keen on your article, since there are not much new data/tips for experienced scientists yet the article would not bring order into rookie’s vision of blunt-end ligations. Again, I am apologetic for a criticism, but I could not withstand a wish to respond to your article.
    Indeed, at the beginning of the 90s, the blunt-end ligation was a solution or even a panacea for many applications or problems associated with Restriction enzymes usage. In the nowadays, at the era of such application as In-Fusion System (Clontech), a Gibson Assembly (NEB), Gateway and GeneArt technology (Life Technologies), LIC, SLIC, PIPE, Transfer PCR, Electra Systems and others, the best way to succeed with blunt-end ligation is not to perform it at all. The techniques listed above may construct whatever one can imagine, thus eliminating every obstacle associated with the blunt-end ligation.
    However, if one needs to clone or subclone his/her blunt-ended PCR products or blunted inserts for sequencing analysis or for mRNA synthesis in-vitro, there are several positive selection vectors for a blunt-end ligation. These vectors come from CloneJet PCR Cloning Kit (Thermo), Zero Blunt TOPO PCR Cloning Kit (Life Technologies) and StabyCloning kit (Delphi Genetics).

    I will agree with you if you tell us that there are still “old-fashioned” scientists who want to introduce their inserts into specific place onto plasmid by blunt-end ligation. Still, there is merely one way to succeed – to use only a dephosphorylated vector and only a phosphorylated on both ends insert and to use PEG at no more than 10% final concentration and at no longer than 1 hour incubation , as Nick wrote previously in his articles about Ligation. As you see, these conditions are obligatory, thus making your Tips 3, 5, 6 and 7 irrelevant.
    Here are some small points of criticism:
    1. In the introduction you wrote “… and with nothing to force directional insertion of the insert, a proportion of colonies will always incorporate the insert in the wrong orientation.” It is not true, since a StabyCloning kit (Delphi Genetics) performs a directional cloning of Blunt-end fragments with >99% clones in correct orientation.

    2. Tip 4: Klenow fragment of DNA polymerase is not so good for blunting of 3′-protruding ends, such as those created by cleavage with PstI or KpnI, because Klenow has a weak 3′-5′ exonuclease activity. The best enzymes for this application are T4 DNA Polymerase, T7 DNA Polymerase or blunting enzyme from CloneJet PCR Cloning Kit (a proprietary of Thermo). These enzymes will fill-in any 5′ protruding end or will remove any 3′ protruding end in 5 to 10 minutes, thus creating blunt ends. However, any thermostable DNA Polymerase with proof-reading activity will perform the job very well.

    Then you wrote: “A better approach is to use an enzyme that produces a blunt end directly (e.g. Sma I, Dpn I, Pvu II, Eco RV) then if you verify the digest, you know that you have a blunt end.” You should correct it and to remove DpnI. DpnI is indeed blunt-end cutter but it is 4 bp cutter that will theoretically cut every DNA after 256 bp! Thus, a 4733 bp plasmid pEGFP-N1digested with DpnI will give 23 fragments making it impossible to proceed with cloning.
    One of the best approaches to produce a blunt ended vector is to perform inverse PCR, as in the Site-Directed Mutagenesis Kits of Thermo or NEB, or as an “entry point” into Gibson Assembly or In-Fusion System. Then use 1-2 ul for subsequent ligation. The huge advantage of this approach is that by combining PCR primers one can “open” his/her vector at any point after any nucleotide not depending on the presence of specific restriction enzyme site!

    3. Tip 6: ” …and phosphorylate the insert.” The easiest way to phosphorylate the insert that is a PCR product, is not to phosphorylate it at all, just to use phosphorylated PCR primers. Any other blunted insert prepared by digest or by filling-in, will contain phosphates on its ends. Alternatively, for blunting as well as for the fragment’s end phosphorylation, I suggest using Blunting Kits from NEB or Thermo. Actually, these Kits are designed for blunt-end cloning of anything.

    4. Tip 9: “Persevere!.. to a point. If it doesn’t work first time, try and try again.”
    Actually, “try and try again” tip was a trigger to respond to the article. It is a bad tip. If nothing was successful in ligation and transformation, then one should stop and perform a little “brainstorming”.

    I suggest very simple algorithm to solve the problem. First, run the rest of the ligation reaction mix onto gel agarose. See page 294 on 2013 Thermo Catalogue for the picture that should be obtained (it’s a good idea to add SDS to the Loading Dye). If ligase reaction performed well, then the problem is either with competent cell (that began to be a little bit “impotent”) or with wrong choice/preparation of LB dishes. Otherwise, if agarose gel revealed that ligation reaction did not work, then ligase may be dead or DNA fragments might contain impurities that just killed the ligase. Simply perform a Control reaction for T4 DNA Ligase activity – just add 1 ul Ligase to Lambda DNA digested with HindIII or BstEII. Incubate 5 minutes “on the bench” and run agarose gel. See Thermo Catologue (2013, page295) or old NEB Catalogues for what that should be obtained – “Jacob’s Ladder” becomes a bulk of DNA fragments. If ligase is alive, then just add your DNA fragments to the same assay to check if they contain impurities which interfere with the Ligation reaction.

    Sorry again for being a “nudnik “,
    Yevgeny Berdichevsky
    Core Facility
    Faculty of Medicine
    Tel-Aviv University

Leave a Comment

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Scroll To Top
Share via
Copy link