PCR with one primer

[exploresci]

This may sound a bit silly but I wonder if you run PCR using only one primer and run the PCR product on agarose gel, what will you see?

[exploresci]

The reason I am asking this question is Figure-1 in this paper, http://www.ncbi.nlm.nih.gov/pmc/articles/PMC385241/

In megaprimer PCR, 'first PCR product' is used as primer for second round of PCR but the 'first PCR product' will have both the strands.

Now if in second round of PCR,I add the megaprimer (2 starnds of DNA as primer) plus flanking primer, two starnds will be amplifying by approaching towards each other on complimentary sides but what about the one strand of megaprimer which will be amplifying in one direction (probably not exponentially)? Because of this one primer extension, will you see various lengths of DNA on the agarose gel in addition to mutated PCR product?

[mchlbrmn]

The single antisense "primer" has no partner, so it can't make a discrete size product, unless the 3' (end of the primer) sequence is repeated somewhere upstream. If the template is circular plasmid, I suppose the two mega primers could prime all the way around the plasmid, if the extension time is sufficient. This would make something large that would be removed by a gel purification. If the PCR product is not purified, I suppose it could reproduce more plasmid (with nicks at the 5' end of each primer) that could add to the original template plasmid and tansform bacteria, if that's the next step. If the template is a DNA fragment, it would linearly produce a single stranded fragment of a certain length, which I think might make a more fuzzy band because of various conformations the single strands might fold into. If the template is genomic, it would go and go and linearly polymerize big stuff of no fixed length, so a faint smear or some band at the maximum size the gel resolves might be seen. But, since it's linear, if the template is genomic, which is very complex, I think this would be a very small amount.
Actually, I suppose any PCR product is also two "megaprimers", so if this were a big problem, it would be a problem in conventional PCR. I suppose any time I PCR a large gene from a plasmid, I'm also increasing the amount of template. I just looked back in my notebook, and when I PCRed some 4-5kb amplicons out of plasmid, I can't see a discrete template band on a preparative gel of the PCR product. Maybe it takes a longer extension time, or an extension temp a bit lower to prevent strand displacement when after it polymerizes on full time around (?).

[exploresci]

Yes, it is so true that the two strands in megaprimer are like the PCR product in conventional PCR.

If forward primer is mixed with megaprimer (essentially two stands of PCR product) in second round of PCR for mutagenesis, only fresh forward primer and reverse primer from megaprimer will make the amplicon exponentially. Whereas, even if the forward primer which has come from 'megaprimer' extends during PCR cycles, it will be linear amplification because there will be no new complementary strands for it apart from antisense starnd of template DNA. Therefore, the variuous single stranded DNA sizes made by 'forward primer of megaprimer' will not be enough to be seen on the agarose gel, though they wil be present in the reaction mixture. Am I getting it right?

[mchlbrmn]

Yes, I think so.
They are only amplified linearly so there won't be much unless the template is cloned DNA (the specific template is then present in large amounts). If the template is genomic DNA, the sizes will be so varied that it will disappear into a smear, or up at sizes larger than the gel can resolve.

However, I think that if the template were a fragment of cloned DNA, for example .1 ug of a DNA strand that ends .3 kb from the "primer", then after cycling you would generate a large amount of single stranded DNA 300bp + the primer size. So in that specific case you might actually generate something you could see on the gel. As I said earlier, it would be single standed, so it would run differently than the size marker ladder, and might be fuzzy, or have multiple bands if the DNA folded into secondary structures.

I'm not sure why more plasmid sized product isn't generated on ciruclar plasmid template, such as when I PCR a large gene on the plasmid. Maybe the extension time isn't enough, although at 5kb that's a large % of the plasmid itself and the PCR product was generated. It may also be that when I do mutagenesis with two primers complementary to each other with the mutation in the center, and "PCR" (not really since it is linear) I think I use a lower extension temperature. I think I use 68C instead of 72. I think this was to prevent strand displacement. With stand displacement the strand extension wouldn't stop at a certain spot where it hits the annealed primer, but could continue around and around the plasmid, and would not make a specific size band. But, I'm not sure 4 degrees would make that much difference. Maybe it's just not very efficient making something so big, and with only linear amplification it's not enough to see.

Whenever any PCR is done, both strands of the PCR product may act as primers in future cycles. The reason a distinct PCR product band is seen is because so much product is generated by the geometric amplification, that it soon exceeds the amount of all other homologous template present, and finds nowhere else to prime. (I'm probably restating what you or I have already said.) In the background of PCR on any larger template, there must be some amount of larger strands generated.