In trouble!

[dttom]

I am contructing a plasmid by inserting a PCR product into a vector. After I did the PCR, I checked the product by running a gel, no problem, the product is then enzymatically cut and gel-ran to remove small fragment, no problem was revealed by the gel. Then I enzymatically cut open the vector, remove the small fragment by gel-run (and check the size, no problem), extract the vector backbone. Next I ligated the PCR product (after enzyme treatment) to the linearised vector by one-end ligation, made blunt the one-end ligated product, ligated the blunt end with ligase to close up the plasmid. After this I transformed the plasmid by electroporation into E.coli. My vector has an ampicillin resistant gene and I spread the transformed bacteria onto an Amp plate; incubated overnight, I picked a single colony and did a liquid culture to grow up the bacteria (supposed to be containing the desired plasmid) followed by plasmid extraction by commercial kits. To those plasmid extract, I performed a single enzyme cut, and gel-run it to look for desired band which is ~7kbp. But I got two bands of 2k and 10k.

 

I'm regretful for not checking the one-end ligated product by running a gel, so it is not promising to have the insert attached to the vector. Even taking this into consideration, empty vector alone should give a band of 4.5k but was not observed in my gel. And I grew it on Amp plate, the competent cells should not have Amp resistance, the option of being contaminated by another bacteria is not so probable...

 

Hope someone could help indicate what problem may have occurred. Thanks.

[CharonY]

To clarify, you got a band at 2k and at 10k (i.e. estimated total size of 12k?). It sounds to me that you may have ligated several constructs together, which would explain two cutting sites instead of one.

[dttom]

Yes the two bands are 2k and 10k. But I'm not sure if it is due to ligated construct, as I also ran the plasmid extract before enzyme cut, there were also two bands, both exceed 10k (maybe owing to circular alignment). The enzyme I used cut only at the very beginning of the vector, no site available in insert, vector I used is ~4.5k, if it is due to ligated plasmid there should be a band of so, but there was none.

[dttom]

The best explanation I could think of is the insert (I) and vector (V) align in an arrangement of (V/I/V|circular), V/V junction ligated in tail-to-tail alignment, so that the enzyme which cut at the very beginning of the vector release an insert (should be 2.5k) in 2k band and vector dimer (should be 9k) in 10k band.

[CharonY]

This is basically what I was thinking. Normally it is best to run each step in a gel, just as a control. However, I am not sure what you mean with:

I also ran the plasmid extract before enzyme cut, there were also two bands, both exceed 10k

 

So your uncut plasmid extract yielded several bands all of which higher than the target size? Of course uncut plasmids run differently, but the CCC DNA should actually be faster than the linear form.

[dttom]

Yes, you got what I mean, the two bands are both higher than target. And CCC DNA, occurring as supercoil, should be faster than linear forms. But if considering them as open circular form, so that they ran slower, the hypothesis of (V/I/V) could explain that, just adding another structure of (V/V) to account for the band near to 10k (two bands exceeding 10k, one near band to and one far band from 10k), upon single enzyme cut it is released to 10k band.

 

I suspect that the problem is I used too much vector or the concentration of reactant DNA is too high to encourage intermolecular ligation rather than intramolecular one.