Not sure about PCR.  You'd need to know where your plasmid integrated to design a primer that can show you the junction. 

More typically I would use Southern Blot for this.  Use a restriction site that cuts once in your plasmid, and a probe against a plasmid sequence.  You'll see the exact size of your plasmid only with non-integrated, free circles.  You probably won't see the integrated DNA at all, because it has all different sizes depending on where it lands.  But you can get the total copy number by qPCR, and then calculate a % integrated. 

If you perform FISH (fluorescence In situ Hybridisation) on fixed transfected cells, with whole endogenous chromosome paint with one color, and the plasmid probe label with a different color, you should be able to distinguish the extra chromosomal (single color) from integrated (dual color),

@ Lingli Li

FISH will only work if the plasmid has a reasonable size.

@Sabine Strehl  

you can detect miRNAs using FISH, Why not a plasmid (several times longer)?

I agree that you can do miRNA FISH, however, not with the usual "home-brew" nick-translation biotin-labeling FISH protocols.

You can use 'Alu-PCR' if your cells are human. If they are mouse there is something similar but different primer sequence. Alu is a retrotransposon element that is interspersed throughout human genome approximately every 4kb. Your first PCR would amplify Alu to plasmid. Your second nested PCR would amplify plasmid only that was amplified during 1st round. Alternative methods also possible, see ref 2 for HIV-1 integration analysis.

Article Quantification of HFV-integrated DNA in human cells by Alu-L...

Article Evaluation of PCR-based methods for the quantitation of inte...

Article Alu Polymerase Chain Reaction: A Method for Rapid Isolation ...

You can also use Hirt methods - SDS and salt precipitates large chromsomal DNA leaving plasmid in the supernatant. but this will only give you an idea for majority as there will be contamination into both samples, so if you are testing for rare events you need PCR.

Article Selective extraction of polyoma DNA from infected mouse cell culture

Thanks for all your responses!

John, I should have mentioned my plasmid has already been linearised for transfection. How would you go about calculating % integrated? 

Lingli, my plasmid is 5kb. Is that large enough?

Nathan, I am using Chinese Hamster Ovary (CHO) cells. The genome is not as well defined as with other organisms, but I believe that using the mouse as a reference could yield some repetitive sequences for this purpose, such as B1 SINE. Do you think that carrying out the initial enrichment PCRs may distort carrying out a qPCR later on? 

I'm not sure a chemical extraction method would work. As you say, there will be contaminants and integration is a rare phenomenon.

Many Thanks

Joe

You can try qPCR twice- once with the entire genomic extraction and once with DNA that is gel extracted for only the long DNA band (basically the CHO genomic DNA). Since your plasmid is only 5kb, extrachromosomal DNA will be much lower on the gel. If you control the amount of loading DNA, this will give you an approximate measure of integration.

When you lyse and wash your cells  with buffers containing non-ionic detergents (eg Tween 20 or TX100), the nuclei stay intact and you should be able to wash away cytoplasmic DNA (i.e the majority of the plasmids, if not all). Maybe that's sufficient for a  pre-selection of single cell derived clones.

Does transfected DNA make into the nucleus in significant amounts and stay there (i.e. gets back there during cell divisions (without integrating) after re-formation of the nuclear membrane? What actually is the (molecular) mechanism supporting/performing the integration?

Wolfgang, I have found that the plasmid is entering the nucleus upon electroporation. The amount of plasmid in the nucleus will gradually be diluted out with cell division, but I assume it will be relatively stable in that environment, so extrachromosomal plasmid DNA is likely to be present for a while. The mechanism for integration is not clear, but the plasmid integrates randomly and integration is rare, so our working hypothesis is that the plasmid integrates into a transient, naturally occurring double stranded break in genomic DNA.

@Sabine Strehl

So, don't use the "home-brew" nick-translation biotin-labeling FISH protocol. Problem solved.

Hi Joe,

I wouldn't use the mouse as a reference because these repeat sequences aren't conserved between species. In fact I've just found this publication which might be your answer?

Otherwise, what about sucrose density gradient like the old method of purifying plasmid DNA? Get your genomic prep, sucrose gradient it, plasmid is denser than genomic. Or good old Southern blot if you can grow your suspect cells up a bit before DNA prep?

Btw, are you testing just because you think it is interesting or it is genuinely important for you to know? I think your explanation is very reasonable, it is known that if you put enough DNA cells there will be some random low level recombination. Gene therapy with AAV or plasmid sometimes shows long-term expression in dividing populations because of this. I'm just pointing out that this isn't too unexpected so might not be worth trying to establish all these assays to prove something already known to occur. Of course I don't know your intentions and you may have a very good reason to pursue it further.

Article The Chinese hamster alu-equivalent sequence: A conserved, hi...

Hi Joe,

yes, I know a qPCR method which can distinguish between integrated and extrachromosomal plasmid DNA.

Briefly, make a tagged gDNA library with RE 1st, in which all gDNA fragment ends were tagged with a universal dsDNA tag !!!. 2nd, run a qPCR with a plasmid specific primer and a universal primer to specific amplify integrated plasmid DNA,, in which neither extrachromosomal plasmid DNA nor gDNA fragment can be amplified !!!

I already developed the whole procedure. If you interested the details, please contact me.

Best regards

Dacheng

Hi Dacheng,

This sounds great. I have messaged you for further details.

Thanks

Joe

Hi Joe,

I can send you a kit with protocols, in which all reagents are included to make the tagged gDNA library, but the plasmid specific primer for qPCR need to be designed on your plasmid sequence.

This is a new biotech and the patent is pending. It is specific designed to localize exogenous DNA insertion site or endogenous transposon jumping in whole genome. This kit was commercialized already but still not officially put into market yet, so I would like give it to you for free trial.

best wishes

Dacheng

Dacheng, the protocol you describe has already been published and is reviewed in one of the references I have already added. I include reference to the original paper describing (what I think is) the first publication of this protocol which is commonly called "linker primer PCR or LP-PCR". You are not an author for this paper nor are you in the same research group, so I'm not sure you can claim it is your work to "have developed the whole procedure"? The lab I work in, and others, also use this technique coupled to high throughput sequencing. Others, and perhaps you, have established it in their own labs, but have not developed the whole procedure, which instead should presumably, as first author, be credited to N. Vandegraaff.

Article Kinetics of Human Immunodeficiency Virus Type 1 (HIV) DNA In...

Nathan,

Thank you for your response. I am keen to use a qPCR method such as those yourself and Dacheng have mentioned, whether it be Linker or repeat element-based. 

I do not have much technical expertise in PCR in general, so I have a technical question about using one of these techniques in my work. After transfection the number of extrachromosomal plasmid copies will vastly outnumber integrated copies. I use 50 ug of DNA per transfection (trillions of copies). Clearly, not all of this will successfully get into the cell, but numbers in the billions of copies may well do. So my question is, even with the PCR enrichment of my genome-integrated plasmid, will there still not be a large proportion of my sample consisting of extrachromosomal plasmid? If so, wouldn't this distort qPCR?

Many Thanks

Joe

Hi Joe,

Yes the extra plasmid makes it more difficult and in your nested you are likely to have carry over plasmid from first round, and 1st round linear amplification of all plasmid (vs exponential for integrated). I use alu-pcr for retrovirus integration analysis which is produced by transfection, and the carry-over plasmid is a pain and reduced sensitivity a lot, but the assay still works (you will have far more to deal with). In alu pcr, you control for this by running the 1st round twice, 1 rxn with alu primers, 1 rxn without. Then, you do 'ct with' minus 'ct without'. If you have integrated copies, you will have a negative number (without had less copies than with).

the ligation presumably could be the same (I haven't used it myself).... with vs without linker primer.

The best way to reduce plasmid is to passage cells as much as your experiment allows. If you can't do that you will likely need to use Hirt fractionationation of your genomic DNA first to remove the bulk of unintegrated plasmid. At least for Hirt, it is a pretty simple procedure, you add a load of SDS and salt and the high mW genomic DNA precipitates overnight on ice/cold room, the low mW plasmid doesn't. It isn't perfect but should reduce your plasmid to levels that allow your PCR to work.

my approach would be:

transfect, grow, passage >3 times if possible - the more the merrier.

genomic prep. test alu-pcr (or LP-PCR if you prefer). if too strong signal in control reaction (ct below ~28), Hirt fractionate the DNA and try alu-pcr again on the cleaned up genomic. if that works you can skip straight to Hirt directly from your cells (no genomic prep in between).

Hi Nathan,

I did not publish my protocol yet. What I mentioned above is not LP-PCR even some step sound like similar between LP-PCR and my method. 

As I know, LP-PCR could amplify many non-specific products and lost many insertions, so it can not be used to quantitate the insertions. 

OK Dacheng, I look forward to reading the publication in the future and understanding what the difference is. Good luck!

Hi Joe, 

"So my question is, even with the PCR enrichment of my genome-integrated plasmid, will there still not be a large proportion of my sample consisting of extrachromosomal plasmid?"----Yes, You are right! So you have to use a method that can separately amplify the two type of plasmid with same sample i.e. my method. Hope you can find another one. If not, contact me again.

Good luck