I think that there are 3 factors contributing. Firstly the primer dimer proper is small so there is much thermal diffusion when running the gel making the PD band very wide so looking somewhat larger than it really is. Secondly there is usually a lot of it although it seems faint because it traps very little ETBR. Because there is a lot of it there is a tendency for the dimer to bind to other dimers and elongate as the amount of product overtakes the amount of free primer so it does get longer. Lastly I think that there can be an effect when the polymerase is getting used up for the melted PD to reanneal with single stranded primer forming a single and double stranded product which will run at a larger size than the double stranded dimer as it has a different conformation again broadening the band in the larger product direction. I have never sequenced PD but when there is PD present the sequencing electropherogram is very short with a sharp drop off of signal intensity  with a very short strong signal dropping rapidly to zero as the sequencing primer and reagents get rapidly used up.

PD could be directly sequenced using pyrosequencing or from clones but is generally too short to get good sequence using Sanger as the sequencing primer and the next 40 unreadable sequence will usually just give you the end of the reverse primer

Hello Xu,

Your primer dimer seems to be of 50 bp. To avoid primer dimer, you may use low amount of primer such as 2.5 pico mole for your PCR reaction. Of course, you may go ahead sequencing it. Best wishes. 

I think that there are 3 factors contributing. Firstly the primer dimer proper is small so there is much thermal diffusion when running the gel making the PD band very wide so looking somewhat larger than it really is. Secondly there is usually a lot of it although it seems faint because it traps very little ETBR. Because there is a lot of it there is a tendency for the dimer to bind to other dimers and elongate as the amount of product overtakes the amount of free primer so it does get longer. Lastly I think that there can be an effect when the polymerase is getting used up for the melted PD to reanneal with single stranded primer forming a single and double stranded product which will run at a larger size than the double stranded dimer as it has a different conformation again broadening the band in the larger product direction. I have never sequenced PD but when there is PD present the sequencing electropherogram is very short with a sharp drop off of signal intensity  with a very short strong signal dropping rapidly to zero as the sequencing primer and reagents get rapidly used up.

PD could be directly sequenced using pyrosequencing or from clones but is generally too short to get good sequence using Sanger as the sequencing primer and the next 40 unreadable sequence will usually just give you the end of the reverse primer

Thanks Paul Rutland and Sharmin Siddique Bhuiyan, nice comments, to avoid primer dimer, you may use low amount of primer for your PCR reaction.

dear Yi Xu

I think as Paul Rutland

I agree with Paul Rutland. 

If your primers are 25 nucleotides long, the primer dimer can not be more than 40 bp as the dimerization needs at least 10 to 15 base-pairing to initiate the reaction. Check the primer dimer formation in Oligo calculator or any other software. You can also avoid primer dimer formation by increasing your annealing temperature by 2-5 degree C.

I agree with paul Rutland, also the primer dimer not necessary mean two primers full length, may be your primer dimer is 40 or may be more than 70 if there is tetra primer or more depend on its compatibility   

What is usually called primer dimers is some material with a lower electrophoretic mobility than the strict primer dimer, as you observed. This material forms a diffuse brilliant spot rather than a defined band as PCR products do. This material can contain primer aggregates (not necessarily dimers) and elongation products formed by the polymerase from them.  Thats why it appears to have a size different from the theoretical one for a dimer. If you dont want to see them, use less primers in your reaction. Actually sometimes they are much more visible in the negative controls than in the positive samples due to primer consumption in the latter. For many applications, they dont represent a problem.

I share the comments given above. Primer dimers are typically seen in ethidium bromide-stained gels as a 30-50 bp or smear of moderate to high intensity. Physico-chemical optimization of the PCR system may help to prevent primer dimer formation. As mentioned above, decreasing the concentration of primers often minimizes formation of primer dimers.

Thank you all. Your comments are really insightful. I just sent the low PCR band (100bp) for sequencing and found there are some other sequences inside the F/R primer. These "short" sequences are just cloned from the 5' and 3'-end of my cDNA template. The predicted size of PCR product should be 1kb. Splicing? Weird?      

But did you only get this small band supposed to correspond to primer dimers or the expected product was also obtained?

I guess that if your ''primer dimers'' band contains something coming from the template its not seen in the negative control reaction without template, is that right? If so, this is not the classic primer dimer band we were talking about, that tends to be stronger in the negative control reactions.

Yes, primer-dimers were sequenced. 

Nucleic Acids Res. 1997 Aug 15;25(16):3235-41.

The elimination of primer-dimer accumulation in PCR.

Brownie J1, Shawcross S, Theaker J, Whitcombe D, Ferrie R, Newton C, Little S.

as your both primer is around 25 bp so the primer dimer will be 40-50bp, to reduces the primer dimer you can use lower amount of primer, hope 1 microlitre of both primer will be enough to reduces your primer dimer. but about the pcr products it depends on the CDS region of your gene where your primer binds.

hope that would be helpful

Hi Yi

As a mode of act you don't have to wander about primer dimers, it happened to me several times to experience having about 100 bp size dimers even if primers were shorter. Consider, by general speaking, there may be several aspecific amplifications from your sample generating these bands.

I suggest you to try to optimize your protocol starting by using annealing temperature at the salt Adjusted primer Tm you have submitting sequences to Oligocalc (http://biotools.nubic.northwestern.edu/OligoCalc.html).

Increase or decrease your protocol annealing temperature if you respectively obtain too much aspecific amplifications or if does not amplify at all.

You will see what happen.

I hope it helps.

AMC

I am finding the band near 100 bp, is it primer dimer. the desired gene size is around 300 bp.Anyone please clarify?

You can also try using a Hot-start enzyme and see if you get rid of the 100 bp band, which as pointed out may likely be a result of non-specific amplification rather than a primer dimer.

As discussed above, the so called primer dimers do not migrate exactly as expected from primer size. They are usualy a bit larger, perhaps because of the presence of more than two primer molecules and also of small amplification products if one oligonucleotide primes the synthesis on another oligonucleotide that functions as a template. Are them present in the control reaction without your target template? If so, they are primer-related artifacts, which usually are not relevant for the final result. If not, they are probably non-specific amplification products, and their importance depends on the relative inensity ecompared to the desired amplification band.