I think ligation has been a common problem for years now. We had a similar problem a few years ago and came to the conclusion that we were getting low yields from gel isolation of DNA and/or modification of sticky ends during gel electrophoresis. We came up with a method for moving inserts between vectors without the use of gels Because it is all done in a single tube there is no issue of DNA yields and we have had  100% success.  It works best when the two vectors have different antibiotics selection but it is possible with the same or if one vector has two antibiotic selections but is a bit more laborious. The method is not only more efficient it is very fast so we dont use the usual methods anymore. You can download the paper outlining the method from my researchgate pages.

https://www.researchgate.net/publication/47680406_STRU-cloning_a_fast_inexpensive_and_efficient_cloning_procedure_applicable_to_both_small_scale_and_structural_genomics_size_cloning

Article STRU-Cloning: A Fast, Inexpensive and Efficient Cloning Proc...

From your 10 ul ligation mix, can you just use much less, say 1uL instead of 5? When you use too much ligation mix your transformation can be overwhelming for some reason. Try old ligations if you still have left overs. -   The other thing to confirm is whether getting out the insert from your second vector is working. XhoI sometimes could be methylated and thus cannot be cut.

After the ligation reaction you can check if the T4 ligase was able to ligate your DNA fragments. Here you can find information on how this is done:

http://www.thermoscientificbio.com/whats-new-t4-dna-ligase/

If this experiment shows you that you have a positive ligation reaction then you can rule out the possibility that your ligase failed. It might at least help you coming closer to solving your problem...

Good luck!

To check your ligation, try a PCR with flanking vector primers (eg sequencing primers on plasmid 1). Remove an aliquot of your lign eg 1ul, dilute 1/10 and use 1ul for PCR. If your insert has ligated properly, you will amplify it up in relatively few cycles (usually

I think ligation has been a common problem for years now. We had a similar problem a few years ago and came to the conclusion that we were getting low yields from gel isolation of DNA and/or modification of sticky ends during gel electrophoresis. We came up with a method for moving inserts between vectors without the use of gels Because it is all done in a single tube there is no issue of DNA yields and we have had  100% success.  It works best when the two vectors have different antibiotics selection but it is possible with the same or if one vector has two antibiotic selections but is a bit more laborious. The method is not only more efficient it is very fast so we dont use the usual methods anymore. You can download the paper outlining the method from my researchgate pages.

https://www.researchgate.net/publication/47680406_STRU-cloning_a_fast_inexpensive_and_efficient_cloning_procedure_applicable_to_both_small_scale_and_structural_genomics_size_cloning

Article STRU-Cloning: A Fast, Inexpensive and Efficient Cloning Proc...

Thanks for all the answers and suggestions!!!

First, yes, I have sequenced my vector 2 (1.8kb insert from vector 1 + pBS vector 2) to comprobate the insert and the mutation after PCR. When I digested with Xho1, I can see my band so Xho1 cuts perfectly both in vector 1 and in vector 2+insert from vector 1.

It is important to me to have the insert in my vector 1 because i want to restore a protein, so it is a bit difficult to introduce/lose any sequence of the insert.

I am going to try to amplify my insert by PCR to avoid gel purification and transform

Hi Eva

 I am a little confused why you are cutting with just XhoI as you would need two XhoI sites at each end of your insert which you want to mutate. For mutagenesis it probably makes no difference doing it this way but going back to vector 1 with XhoI your insert will go in  in a forward or backward direction an I guess only the forward will express correctly.  Secondly we do not use the same vector 1 + insert  when moving back from vector 3. Instead we make a large supply of vector 1 which only has the fragment with  restriction sites, in other words the vector you bought from the manufacturer. So the tiny insert made from this vector readily passes through the spin tube and is lost so the only insert present will be the one coming from the pBluscript vector. Also doing it this way you can guarantee that there will not be any original insert present in this procedure only the mutated.  To summarize you made

     V1-insert     +   pBlue     > XhoI>   V1     +    pBlue-insert

now you need to do

     pBlue-insert   +  V1    >XhoI>     V1-insert   + pBlue

unfortunate     cutting with XhoI will make both V1-insert and  V1-tresni  so you really should use two restriction sites to conserve directionality in V1  

Sorry Kristin, maybe I explain wrong! I actually did a gel purification step to remove my unwanted insert 1 and then I dephophorylated the vector (followed by another purification as the protocol said). Then I used it to ligation.

Maybe, should I religate my vector 1, do a maxiprep and then digest and dephophorylate it? I could avoid the step of gel purification vector 1 by this way and it could help to ligation...

What do you think?

Thanks again!

Hello Eva,

As you are using single restriction enzyme (Xho1) for cloning; there is lot of chance to self ligate the molecules due to the complimentary between 5' & 3' flanking of same molecules ( may form monomeric linear to circular [or] concatemeric of linear molecules)

As your vector dephosphorylated, it may not self-ligate; but your insert will definitely form concatemers/self-ligation due to the complimentary between 5' & 3' Xho1 flankings.

In this case two phosphodiester bonds will take place in between two insert molecules but in the case of "vector + Insert" it will be limited to one.

And moreover the accessibility of sticky ends more in between same molecule than different molecules.

So by considering the above facts:

1) Don't  dephosphorylateyour vector, as well as your insert also

2) Before adding T4 ligase in your reaction, heat your Vector & Insert DNA molecules at 450C for 5min to open the loop formation of Xho1 flanking sequence (after digestion Xho1 leaves "TCGA" overhang, so loop may form due to complimentary between T->A and C->G)

3) screen more clones for finding positive clone after transformation; because Non-dephosphorylated vector may give some false positive background.

All the best...

Hello everyone!

Your answers are really interesting. Some clarifications here:

I have to use only Xho1 because my gen is too large to use all of it and the part I wanted to mutate is surrounded by these restriction sites (Xho1). There were others of course, but they were also at other part of my gen and in pBluescript, so it wasn't possible to me to use other enzymes. I would have preferred to use 2 enzymes but it was completely impossible.

My question for Miroslav is: (forget about my particular case) Could both vectors ligate each other as they are at the same tube and the resistance let them to grow?

Kristin, I have colonies in which my vector is religated, so the amount of vector it couldn't be a problem..

Suneelshekar, I am going to probe it right now...seems pretty interesting and it may work although I will have to do a bigger screening!

Thanks everyone, you are helping me a lot! If you have more ideas, please tell me :)

I will tell you if any of your ideas finally work!

Eva-

Hi Eva,

I will run ligation on a gel to figure out whether it is a ligation problem or transfection problem.

Lee

Hi Eva

   Yes there is a chance for cross ligation of the two vectors but the probability is low. If you read the paper for the example in figure two 4 out of 5 colonies selected were OK and one crosslinked of two vectors.  That is why we normally select a few colonies at random after ligation  and plating out followed by plasmid isolation to check in gels. But crucially you dont need to isolate these  from gels. All of this is a couple of  days work and most of that is waiting for colonies to grow then growing selected ones in 15ml tubes..

Hi Eva,

Just a quick clarification, when you tried the non phosphorylated vector and got self ligationws was it gel purified ?  I ask as we had a lot of problems with a pHD students cloning, and discovered that the agaorse we were using was interfering with the ligation.  I always use low melting point agarose for gel purification irrespective of what method I use to purify the bands.  Perhaps a quick ethanol precipitation of the purified vector/insert will also help with purity.  What buffer is the DNA resuspended in as if you are adding  5-10 ul to a 10 ul reaction, it would be best if it was just in water to avoid perturbing ligase buffer concentrations.   One suggestion that will help both with the gel purifcation and will prevent the vector with the insert in it from perpetuating if it is present, would be to cut the insert vector with an enzyme that doesnt appear in the insert, dephosphorylate and purify (if necessary), then revove the desired insert with Xho1.  Also If you tried a ligation into any other Xho1 cut vector it would let you know if the insert or the vector is causing the problem.  If you are always gel purifying try a different agarose, as there does not seem to be that much wrong with your approach so far and it took us a long time to find out our agarose was causing the problem.

Hope this helps and good luck with it

David

I would switch to Shrimp Alkaline Phosphatase (https://www.neb.com/products/m0371-shrimp-alkaline-phosphatase-rsap) because it can be 100% heat inactivated. The CIAP you are using could be carried over to the ligation reaction and dephosphorylate the insert. That would explain your results.

I would continue with dephosphorylating your plasmid but as Gary has stated you either need to clean up CIAP treated prep after use or use another phosphatase. 

The insert will self associate: you can check this quite easily by running a sample on an agarose gel. I would  carry ligation out at 37C for 2hrs with 2x more ligase. I would also try using a much lower amount of insert than you have been using to reduce competition for sites. Always use fresh aliquot of ligase buffer, ATP goes off over repeated freeze thaw cycles.

If you have a lot of colonies I would PCR screen your colonies for inserts using one vector and one insert primer pair to speed up process of selection: this way you can screen hundreds of colonies at a time: Pick colony onto grid plate and grow o/n the remainder is placed in a PCR reaction: off you go.

good luck

Thanks everyone for your advice and suggestions. I have tried some of them but i didn't have any success. For example, Suneelshekar's suggestion was heating heat my Vector & Insert DNA molecules at 45ºC for 5min before ligation, but i got two colonies and they were religation of the vector. 

Now, I am cleaning up CIAP treated vector before ligation and I am going to use electrocompetent DH10B cells. If it doesn't work I will finally buy Shrimp Alkaline Phosphatase, I really think that my problem is the step of dephophorilation. 

By the way, could anybody recommend me a low-melting agarose? Which one are you using David Andrew? The answer for your question is yes, I did gel purification of my non-dephophorilated vector and I got religations. 

Thanks everyone, I hope we will find my particular solution! :)

Eva-

Hi Eva,

I generally use seaPlaque GTG, however this is very old, and probably replaced by another version I think Lonza sell seaplaque now, or ultrapure from life technologies will do the same job.  Any agarose certified for in gel reactions will be fine.  I use antarctic phosphatase, as a heat inactivatable phosphatase, and have also found PCR cleanup kits from qiagen good for removing phosphatase and the removed DNA, providing the DNA fragment removed is Small (under 50bp).

  I know a lot of people offered lots of advice on this column, regarding ligation conditions and dephosphorylation all of which are very sound advice, but the fact that you can do the first ligation, suggests that your methods are allready sound.  You should be examining what is different between ligation 1 and ligation 2.  The only real difference that I can see are two gel purifications before the second ligation.  Hence my sugestion about the agarose.

You should check out the insert and the vector.  You could quickly make an insert from two oligonucleotides bearing xho1 overhangs on each side. and try to ligate that into vector 1 to check that.  Put a unique restriction site in the middle of the cassette and you have a quick way of checking the transformants.

As for checking out the insert, you could always blunt it and clone it into a sma1 cut vector.  However this will not check the restriction sites totally, as you would only be able to fill in the overhangs.

It does seem a mystery that it does not work in the second ligation, I would be happy to have a quick look at the sequences if you wanted to send some details.

Hope this helps

David

  What is the vector for vector 1, as I know that vector 2 is bluescript.  I dont think this could be too much of a problem though as even if you were getting leaky expression problems as you are only using Xho1 it would just select for inserts in the reverse orientation

Hi Eva,

Sorry for the last bit of my last response, I did mean to delete it, as I didn't think it added anything.  I did mean to say, from the techniques you have employed you do have the resources to try a gibson assembly method for your second ligation.  Have a look at the Gibson assembly kits on NEB, you would need to amplify your mutagenised product with new primers with overhangs that match vector 1, but it would add directionality to the "ligation" of the insert which would save screening time.

It does work really well as I have used it to clone into really large vectors where I simply did not have enough restriction sites to play with.

David

Hi everyone,

Here I am again. I have performed some of your suggestions but I am still stuck.

I have used an ultrapue low melting point agarose from Invitrogen, and I have changed my phosphatase for Shrimp Alkaline Phosphatase (rSAP).

I have performed different protocolos:

a) Xho1 digestion - gel purification - dephosphorylation with rSAP - ligation

b) Xho1 digestion - dephosphorylation with rSAP - ligation (without gel purification)

It seems no changes. My control ligation (without dephosphorylation) is ok because I have religation. I don't know why it is being so difficult to clone my gen.

Some of you suggested me to do a PCR: I have to restore a gen (a protein) so I can not add any pb in my gen, also products from PCR are dephosphorylated, so I would have to phosphorylate my gen, adding one more step to difficult the clonning, isn't it?

As some of you asked, I enclose here a figure to show you my vector so that you can see it and tell me if yoy think there is a ligation / digestion problem.

Thanks for your answers, hope we can figure out my problem!

Eva-

Also I have used electrocompetent cells (DH10a) in a 1mm cuvette!! So transformation is not probably the problem...

Hi Eva

  It is difficult to say what you should do or what is going wrong without detailed knowledge of your vectors.  Have you sequenced your gene insert in pBluescript including the surrounding DNA that will include restriction sites and do you have the DNA sequence of the plasmid you wish to insert into.  If I had these two I could advice you better. If you do not want to give these sequences here I could send you my email.

Hi everyone,

Thanks for all your answers. Finally I got my positive colonies!!! All I had to do was to increase the amount of vector a lot. That was actually my problem with ligations and I improve them with some of your advice, so THANKS!

Now, I want to do a different cloning. I have to cut the vector pBabe with BamHI and EcoRI and perform a ligation with a linker I had cut with the same enzymes. My problem is that I have many colonies in my control ligation (only vector) so I think that the vector is not well digested.

I have done a double digestion with both enzymes (from Promega, buffer E) and then gel purified it.

Also I have tried a BamHI digestion, gel purification and EcoRI digestion.

It is not necessary but I have also dephosphorylated the vector but it seems no difference.

BamHI and EcoRI restriction sites are separated 24pb. 

Anyone has an idea of my problem? 

Thanks again

Eva-

I am glad your cloning worked. The relative amounts of vector/ inserts etc. was the problem we kept coming up against that is why we invented our cloning method i mentioned early on. Because it is all done in one tube it can be guaranteed that no vector can be lost in intervening stages such as gel isolation but it does need different antibiotic selection to work easily.

With your current problem you don't provide enough information to advice you the following is needed

(1)  what are the fragment sizes from pBabe after cutting with BanH1 and EcoR1

(2) how did you get the linker was it from another vector or was it manufacture. If from a vector which one

(3) what is the antibiotic resistance of all vectors.

I assume the new linker is being ligated with BamH1 and EcoR1

Thanks Miroslav for your answer. Here I add some information you asked me about my problem:

1) After cutting with BamH1 and EcoR1, I obtain two fragments: 5145pb and 24pb because the restriction sites are separated only 24pb.

2) The linker was designed and then bought in Sigma. It is cut by BamH1 and EcoR1 to ligate it into pbabe.

3) pbabe is ampicillin resistant

I hope this could clarify something!

Eva-

It should be straight forward. When you say you only get colonies which are  only pBabe how do you know some of them are also pBabe with the new linker. If you ligated with the old linker still present then you will get a mixture. Isolating cut pBabe on gel will get rid of old linker but you don't say if you got any colonies from this after ligation with new linker. If you are not if you did get colonies it could only be with the new linker if you did not get colonies then the manufactured linker did not cut  fully. I can seen this problem of  cutting manufacture DNA for efficient cutting you would need  to manufacture a linker with several more bases beyond the BamH1 and  EcoR1 sites. If you think the pBabe is not being fully cut then you should see on a DNA gel significant amount of circularised vector as well as linear vector. If you only see a linear vector in gels then this cannot be the problem.

If you are only getting colonies in your first experiment (ie cutting and then adding new linker/religating) then it maybe that you are not getting rid of all the old linker and it is religating especially if you dont have enough of new linker.  To ensure this does not happen use a  modified version of our STRUBI-cloning method. After cutting your pBabe vector use something like a Vivaspin® 500 Centrifugal Concentrator with a 100,000MWCO and spin  to concentrate a few times after diluting with buffer. This repeated dilution and concentrating will remove the 24 bp linker. It may also remove the BamH1 and EcoR1 enzymes but I would be cautious and heat inactivate the enzymes, usually 65degC for 5min. This will ensure you have only cut pBabe but none of its linker. You can then introduce the manufactured linker and ligate normally  You may have to play around with the cut vector to new linker ratio to get the best results.  If you dont get colonies this way then I suggest you look at the linker to see if it is being cut proparly the only other problem can be if some of the pBabe is not cut at all.,  The real problem is that you have not selection for the new linker.   The only way of being completely certain that both pBabe and new linker are cut is gel isolation of both but to see if the linker is cut fully i would make a larger bit of DNA maybe 500 bp with linker in the middle so that when it is fully cut you will see approx 350bp fragments but no 599 bp fragment.

These are some thoughts I have but you can see there are a few possibilities about what maybe going wrong.

Sorry for my last answer. I read through it after it was sent and it was a little scrambled. I have corrected the mistakes below.

It should be straight forward. When you say you only get colonies which are  only pBabe how do you know some of them are not also pBabe with a new linker. If you ligated with the old linker still present then you will get a mixture pBabe with either linker. As you suggested isolating cut pBabe on gel will get rid of the old linker but you don't say if you got any colonies from this method after ligation with new linker. If you did get colonies it could only be with the new linker if you did not get colonies then the manufactured linker did not cut  fully and there was not enough to ligate.

I have seen this problem of  cutting small DNA before because for efficient cutting you need  to manufacture a linker with several more bases beyond the BamH1 and  EcoR1 sites.#

If you think the pBabe is not being fully cut then you should see on a DNA gel significant amount of circularized vector as well as linear vector. If you only see a linear vector in then this cannot be the problem.

If you are only getting colonies in your first experiment (ie cutting and then adding new linker/religating) then it maybe that you are not getting rid of all the old linker and it is religating especially if you dont have enough of new linker because the balance will be tipped in favour of the old linker religating.  To ensure this does not happen use a  modified version of our STRUBI-cloning method. After cutting your pBabe vector use something like a Vivaspin® 500 Centrifugal Concentrator with a 100,000MWCO and spin  to concentrate after diluting with buffer. This repeated dilution and concentrating will remove the 24 bp linker. It may also remove the BamH1 and EcoR1 enzymes but I would be cautious and heat inactivate the enzymes, usually 65degC for 5min. This will ensure you have only cut pBabe but none of its linker. You can then introduce the manufactured linker and ligate normally  You may have to play around with the cut vector to new linker ratio to get the best results.  If you dont get colonies this way then I suggest you look at the linker to see if it is being cut proparly. The real problem is that you do not have selection for the new linker.   The only way of being completely certain that both pBabe and new linker are cut is through gel isolation of both but to see if the linker is cut fully I would make a larger bit of DNA maybe 500 bp with linker in the middle so that when it is fully cut you will see approx 250bp fragments but no 500 bp fragment.

These are some thoughts I have but you can see there are a few possibilities about what maybe going wrong.

Thanks for all your thoughts! I think the idea of making a larger fragment of DNA with my linker of interest in the middle could be great. 

However, I still think that part of my problem is in fact in the double digestion of pbabe. I cut my vector with both enzymes, gel purified it and perform a ligation:

- One without my linker (control ligation)

- One with my linker (ligation of interes)

I do get colonies in both of them. I have picked some colonies and made a miniprep and I sent it to sequencing. The linker wasn't in the plasmid, it was self-religation of the plasmid. 

So, the problem of my old linker (24pb) is solved thanks to gel purification. I could expect no colonies in my control ligation, or at least less colonies than the ligation of interest. 

If it helps, here is my linker sequence:

Fw: 5' GGCCGGATCCATAAGCGATCGCAATATGAGCGGCCGCTTTAGTAGTTTAAACAAATGAATTCGCCA 3’

Rw: 

5’TGGCGAATTCATTTGTTTAAACTACTAAAGCGGCCGCTCATATTGCGATCGCTTATGGATCCGGCC 3’

The linker contains other site restriction so i could check ligation with a single digestion or by sequencing. 

What do you think?

Thanks

Eva-

You could check other sites for cutting I guess. But  I would first ask which buffer are you using because EcoR1 can have lower cutting efficiency in certain buffers than BamH1. Also you could check the fidelity of each enzyme in case  one has lost its activity a bit.  In other words with the same buffer for each enzyme look to see how many enzyme units you require to linearise the plasmid. Do this for each in turn. It maybe that either you need a different buffer. NEBuffer 3.1 is recommended to cut both 100% I think it is 100mM NaCl

50mM Tris-HCl

10mM MgCl2

100μg/ml BSA

pH 7.9@25°C

So investigate the quality of the enzymes and the correct buffer first. You may have already done this and if they are OK then I am not sure what to suggest.

@ Miroslav Z Papiz   What is the ratio you are using for ligation of cut donor and vector plasmid without gel purification of the insert?

we generally use 10 ul of vector at 0.2 ug/ul and 10 ul of insert vector at 0.45 ug/ul. However tests indicate  it works over a much wider range of ratio Use straight from plasmid miniprep isolation

@ Miroslav Z Papiz Thanks a a lot for your quick answer !