I don't think supercoiling is a likely explanation, because any supercoiling of the starting materials would have been eliminated when they were cut by restriction enzymes, and DNA ligase would not introduce supercoiling. In other words, the assembled plasmid should be in a relaxed state. I think it's more likely that the assembly did not go as planned.

Hi Madhavi,

maybe you could give some more information? (as Adam B Shapiro mentioned I also think that somethink went wrong with your assembly).

What vector did you use for cloning and what is the size of the vector?

What size is your insert?

Which restriction sites are you using?

Do you try to ligate a PCR product (and if yes what was the template for the PCR product - maybe another vector?) or did you cut out the insert from another vector?

It would be also interesting to see a picture of your gel.

Maybe I can come up with an idea with this information (however I cannot guarantee that I can help but maybe another researcher will have an idea with more information).

K Dörthe

Dörthe Andrea Kesper Hi Dorthe,

Thank you so much for your response and for taking the time to ask these helpful questions! Here are more details about my experiment:

I’m working on assembling two synthetic DNA fragments, one is 3769 bp and the other is 3936 bp, so the expected assembled product is around 7.6 kb. Both fragments were synthesized (not PCR products), and I’m assembling them using BbvCI and SfiI restriction sites. There is no vector backbone involved; it’s a fragment assembly.

Interestingly, I came across this paper: “The SfiI Restriction Endonuclease Makes a Four-strand DNA Break at Two Copies of its Recognition Sequence” (Wentzell et al., J. Mol. Biol. 1995). It seems that efficient cleavage by SfiI depends on having two recognition sites in proximity, and cleavage can be inefficient or incomplete when only one site is present per fragment. I think my assembly issue could be related to this unique behavior of SfiI.

At this point, I’m considering switching to a different restriction enzyme to avoid this complication.

Thanks again for your time and suggestions

Adam B Shapiro Thank you so much for the explanation

Hi Madhavi,

I am sorry that I misunderstood your approach. However, I must admit that I never have done an assemby of a plasmid with synthesized DNA products and I never have used one of the restriction enzymes you used, so I am not sure if I am the best person to give you advice.

However, I have looked up the SfiI site and you are correct - the SfiI enzyme need two recognition sequences with 5 bp in between (N) for eficient digestion.

So long as you haven't constructed your plasmid accordingly, it is unlikely that this enzyme will cut efficiently. Check out the NEB site - here you are getting tips for efficient digestion if you are working with this site.

https://www.neb.com/en/tools-and-resources/usage-guidelines/restriction-enzymes-requiring-multi-sites-for-efficient-cleavage

Regarding the supercoiling - this will only occur if you have propagated your plasmid in bacteria after assembly, but from your description it doesn't sound that you have done this, especially as you have mentioned that there is no vector backbone involved.

If I got this wrong at you propagated your plasmid in bacteria supercoiling is a possibilty - in this case I would look up if you have another enzyme with a single restriction site and try to digest the plasmid with this enzyme to see if you can linearize the plasmid.

You might also consider to send the assembled plasmid for sequencing to see if you got is what you wanted to get.

If you don't have the desired product and you don't want to try the recommendations on the NEB site I would indeed recommend to try another restriction enzyme for assembly.

I hope this helps a little bit and maybe some other researcher can give you aditional advice who is more experienced in the kind of experiment you did.

KR Dörthe