Protocol Online logo
Top : New Forum Archives (2009-): : Molecular Cloning

Electroporated plasmid loses the insert - (Jan/20/2012 )

Hey folks, quick question for you.

I have inserted a mutation construct into the pGP704-based plasmid pRE107 (pir dependent replication suicide vector). The plasmid was electroporated into E. coli CC118 lambda pir, selected, purified, sequenced, and found to have been constructed correctly. For conjugation, we use the E. coli strain SM10 lambda pir. I did a plasmid prep on the CC118 culture, verified it to be the correct size, and verified the presence of the insert by PCR. Upon electroporation into SM10, I get multiple colonies on LB-amp, but every one of them that I have screened does not contain the insert as determined by size of the plasmid on gel and PCR using plasmid-specific primers that flank the insert site and insert-specific primers. I am going to submit the plasmid from SM10 for sequencing to see if I can get any clues to what is happening. I also have the E. coli strain S17-1 lambda pir if I need to go into there.

This one boggles me, because I've done a lot of cloning into this vector, and have transformed CC118 and SM10 many times without this occurring. I don't get how a purified plasmid containing an insert would be reverting to empty vector following electroporation. Any thoughts on what might be happening?

I did check the insert for repeats that could possibly lead to it looping out, but nothing significant comes up. It's possible that E. coli could have a similar gene in the genome, but by my understanding, the E. coli strains should not be able to recombine the DNA contained in the plasmids.

I appreciate any thoughts.



Its quite perplexing, SM10 lamda pir is recA negative, so chromosomal recombination is unlikely as you said. One possibility is that your plasmid prep from the CC18 culture is a mix population of two plasmids - one with the insert and one without. In which case, especially if the empty plasmid is predominating this population, you would get this observation upon transformation with an aliquot. Have you checked the plasmid prep from the CC118 culture with plasmid specific primers in addition to the gene specific primers?


Thank you for the reply.

Yes, I did check the prep from CC118. The PCR product was a pure product at 1500 bp, indicating the correct insert, using primers 100-200 bases outside the insertion site. There were no bands, even faintly, in the area of 300-500 bp, which is the expected product from an empty vector. The preps from SM10, however, were purely bands in the 300-500 bp range with nothing close to 1500 bp.

Also I forgot to mention, this is the 3rd electroporation into SM10 that I have done, with all having the same result. The first two were with the same prep, thinking that the first I may have just grabbed the wrong tube from the freezer. The second was from the correct prep from the original CC118 prep that was sequenced, and the third was from a new prep from the glycerol stock made from the original culture.

I fear this may be one of those instances where it's not be worth it to spend the time and money to figure out what is happening, and just move to trying a different conjugation strain, e.g., S17-1. Will be quite disappointing to move to that strain and find the same problem, though.

I will definitely be sequencing the cloning site of the "empty" vector coming out of SM10 to see if there is any insert left or if it is normal and empty pRE107.


Does your insert has homology to the E. coli genome? ...there is also something called recA-independent recombination ...and if your insert is somehow toxic recombination event can rapidly lead to outgrowth of your insert containing plasmid population.



I have not blasted the ORF against E. coli. It is possible it could have a similar gene. What is in the vector, however, is a deletion construct. There is a significant portion of the internal sequence deleted with probably about 500 bp of flanking sequence on either side. To my knowledge, the flanking sequence also does not contain a full gene (the entire sequence is part of an operon) The full size target gene is expected to be a structural component of a virulence secretion system, so I don't believe it would be to toxic to E. coli, particularly as the DNA in the vector a deletion mutant.


virulence secretion system ...sounds like you are talking about cloning sequences of prokaryotic source :)

Why do you think there is likely no homology? ...i would check that first by doing a blastn. What if you are deleting E. coli genes as well using your plasmid ...then i could clearly understand why your insert gets deleted instead of deleting a gene on the chormosome (that maybe is essential).

I will try to summerize your approach far is i have understood it (and please correct me if i'm wrong ...maybe i missed some details):
You are trying to clone a plasmid that you are going to use for deletion studies? ...vector contains homologies to 5' and 3' homologies to the gene you are going to hit in your bug of interest ...true? the middle you have an antibiotic resisitance?

Just a (maybe strange) hypothesis ...what if the two pir+ strains differn in their genotype in a way that one tolerates your insert and the other does not?
Maybe you can use a different pir+ strain like B2155.



I figured the SM10 was derived from a non-virulent isolate of E. coli, so the likelihood of it having DNA homologous to a virulence secretion system is low. That is just a guess, though. I did blast the nucleotide and protein sequences of the target gene against E. coli, and there is not significant match for DNA, but the protein sequence does share homology with a protein in some E. coli isolates. But without knowing the sequence of SM10 specifically, it is difficult to determine if any of the DNA would match. I do know that I have constructed 5 other mutations in genes of this secretion system and none of them have done anything like this. I could try a PCR on SM10 using target gene specific primers to see if anything is amplified.

As for your summary, all is correct except for the resistance gene. The mutation is a markerless deletion. I can put a Km marker in there, and that would help select for the maintenance of the insert in the vector; however, the target gene is in the middle of a large operon, and insertion of the Km marker would likely cause polar effects on the downstream genes.

As for the pir strains, that is possible, and we do have another conjugation strain that is available to use (S17-1).


Got the sequencing back on two plasmids isolated from SM10. Both are not only empty vector, but intact empty vector. Cloning was done with SalI and XbaI, and the fragment that should've been removed by digestion was still there. I can see possibly the culture that was frozen having some contamination with any uncut, empty vector that may have survived the digestion and went into a cell during electroporation. What I don't get, however, is after running the prep on a gel and doing PCR with plasmid-specific primers, I get no indication of any contamination. If the contamination is low enough to not be seen on a gel or by PCR, it is odd that it is so well represented in the transformed colonies.

Anyway, I have prepped some S17-1 E. coli competent cells and will do an electroporation with S17-1 and SM10 side-by-side to see if the lose of the insert is strain-specific.


i would suggest doing all the necessary controls next time you do an ligation experiment, meaning
1) vector+insert+ligase
2) Cut vector only
3) Cut vector +Ligase
4) Insert only

to prevent misinterpretation of results. This gives you a feeling for the amount of false positive clones resulting from uncut vector or religation of the cut vector.