These are single-stranded products that from secondary structures.  A few years ago, I excised and sequenced such bands and they had exactly the same sequence AND LENGTH as the primary product.

We perform EVA Green qPCR and perform them until 10 cylces in plateau phase. Then, in some reactions a new amplification starts to be seen by an steady increase in fluorescence. In corresponding melting curves by a second peak with a higher melting peak occurs.

As we always perform qPCR to 10 cycles plateau we thinkl primers are exhausted or cannot compete for amplification anymore. We know that by less primers lowered plateaus are present, indicating that primers are limiting. So the additonal products (bigger than the amplicon) cannot be (hetero)primerdimers. We think that surplus of melted amplicons not precisely reanneal, but leave sticky ends that are elongated to the 5' ends of the amplicons allowing the rise in ds. DNA binding Eva Green. We are planning to sequence them directly to proove this hypothesis.

Hi Christian

There can be several reasons for this observed behaviour...and since you are not new to qPCR assays...I will belive you have taken care of many things while designing the PCR primers (includding intron-exon organization etc.)....but still I will have some queries and some suggestions

1) What is the nature of gene...single copy or multicopy...does it have mutiple members in the genome

2) If yes, does your primers are designed specifically to amplify a single member

3) Secondary structure is also a possibility (you can check this in denaturing conditions but you maay not get a complete answer)

4) One possibility which we have encountered in the lab and struggled a lot to solve is the 'heteroduplexes' formation between the products formed between the amplicons from different members which are formed during repeated denaturation and reannealing steps....and we could not solve in by cloning and sequencing separate bands.....the actuall differences were very small indels not visible side by side in the lanes ...but when you mixed the products and did analysis...you could reproduce the ghost bands....it may also be true in your case....so clonin and sequencing may give you some idea and as I also mentioned earlier..carefull analayze the nature of the gene...see one of our paper on such a problem.....(of course there are many such paper)

Saini et al (2008) Genet. Res., Camb. (2008), 90, pp. 299–316.

I hope you may get some insights into the problem

bye and all th best

A different secondary structure seems most likely when the sequence is the same, you might use denaturating gels (PAGE) to confirm this.

Other ideas:

A different allele, i.e. the samples with a second band would be heterozygous. But this would cause sequencing problems.

Unspecific amplification (same problems).

Primer dimers (no idea how large your fragments are, primer dimers are usually less than 100 bp).

Edit: too late...

Once it is a specific and quantitative so revise the fragment polymorfisme and the possibility of different allele

Maybe are primers-dimers.

Hi Natacha

These do not look like primer-dimers as they are well above the expected products..primer-dimers are generally below 100 bp and near the bromophenol dye front,,

One possibility is that you are looking at a salt effect from your PCR mixes - as the gel runs, the salt front moves ahead and this may affect the way the product runs i.e you are looking at a salt gradient effect.The sample is homogeneous but the migration is affected. A simple test is to add PCR mix to your markers and run them alongside marker without PCR mix to see if there is a salt effect. When we run Polyacrylamide gels we usually have a stacking gel that concentrates the protein and it becomes well separated from salts. In agarose gels the presence of salts in wells at the start of the run might effect the way the products run - the more sample /salt loaded the greater the effect.

Just conjecture......

Ian

maybe you need to optimize the concentrations of primers. If your all lanes in this picture are similar product, the secondary band may be single strand PCR product.

Re ss PCR bands, - Interesting idea, but if you can clone the upper band then it is not likely to be ss. You could also test this by heating the DNA to 95+ for 3-5 min and then snap cooling then running out on a gel ( + formaldehyde ??) - everything will then be ss with no ds.. On the other hand, do you really want to spend the time and does it really matter ? ....

Ian

Re-optimize the PCR product. Maybe you can check the concentration of dNTP and MgCl2 ...

In my opinion, i´m very agree with the answer from Ajay Saini

Christian, the same thing has happened to me with some PCR products. I agree with you that these bands are not unspecific products. I am sorry but I do not have an explanation for this, but please, tell us what you find out. I would like to know why this happens sometimes.

I think that we are missing some other information to interpret these data. What type of analysis is this? PCR from cDNA or from gDNA? Are you using eukaryotic or prokaryotic samples. If you are dealing with gene expression analysis (so cDNA synthesized from your isolated RNA samples) in eukaryotes, there is a high chance that your primer set is not intron spaning, which can result in the amplification of the corresponding gDNA, especially if the intron is of small size. Your second band is high enough (+/- 250bp), which would indicate that the intron is about 100bp. The sequencing data will definitely help you to resolve this issue. And if that happens to be waht I suggest here, you have to design new primers, with at least one of them that spans two exons. Regards and success!

This happen to us very often as well.

My opinion: these could perfectly be dimeric configurations of your product. Notice that MW is almost 2x that of your product. In the last cooling step of PCR, some cross hybridizations could occur.

In any case these are NOT unspecific products

I agree with the opinion that a "ghost" band could be a dimer of the main product. If the template is genomic DNA it can well be that there are two copies of your target sequence in the genome that differ in the organisation, such as the intron length. In addtion there might be a site specific duplication leading to tandem arrangemet. If so, the polymerase would amplify both monomers and dimers. If the larger "ghost"band is a real problem you may wish to get rid of it by decreasing extension time in your PCR reaction. This would favour amplification of short fragments while amplification of larger fragments will be suppressed..

I experienced such ghost band many times... I don't know its origin, but, when I purified and sequenced the PCR product I was visualizing, the sequencing reaction has always been perfect! Therefore, don't waste time to understand what it could be, probably it's only an artifact, maybe due to agarose low quality, typical of agarose for routine use.

No worry. Try this. I guess, this is the primer sensitivity to the annealing temperature. I do not see your PCR condition here. Anyway, Increase your annealing temperature and then run the gel. Provided prepare the DNA with all newly prepared reagents. Discard all of your old reagents. Let me know this works. IMJ

HI Christian

I also agree with comments of Jesus and Ales....it could be the prefect diemric configurations if tow copies of a gene are arranged 'in tandem arrangemnent' in the genome.....So there are several factors and check them one by one......if finally they are ssame sequences and do not affect your analysis or results in some way ..then as suggested by Nicola...don't worry too much .....WE have seen such results...as a consequnce of tandem gen arranagement during our analysis of 5S gene in plants where not two but several genes are present in tandem...and you actaully see a ladder type of fragments in the gel .....so you also choose appropriate extension time...for your locus

l agree with Jesus and Ajay-perfect dimeric configurations of your product apparently hybridising in the cooling step. The question is maybe how to eliminate that.

Thanks for all the input.

As mentioned, I cloned both fragments which I first extracted independently from one of my gels. I have now sequenced the inserts and the sequence appeared to be identical.

My input is genomic bacterial DNA.

The question is indeed now, how do i reduce or eliminate this. It would be nice to be able to visually show my PCR is specific, besides showing one melting curve (and sequence data). I will play around with extention time of my PCR (now I run 40x 95C for 15s, 58C for 20s, 72C for 30s, primer Tms are 61.7 and 63.0 C).

I have experimented with a PCR product purification spin column and/or 5 min denaturation step (and cooling down) before I load it on the gel. Strangely the denaturation step increased the conc. of the ghost band alot.

Have you check

Magnesium concentration?

What is the gene? Some, such as control region sequences are more prone to forming 2nd'ry structures than others.

If the sequences of long and short fragments are identical it is likely that the product dimerizes during a PCR reaction. One reason could be the presence of subrepeats within the amplicon. You may wish to check for the repeats by dot plots /pairwise comparisons/ at http://bioinfo.lifl.fr/yass/yass.php. YOur observation "Strangely the denaturation step increased the conc. of the ghost band alot" is significant in that the dimer may not be easily eliminated by PCR set up since it can form anytime perhaps even without an elognation step.

How about increasing your annealing temperature and magnesium concentration?

I am agree with Marzan Mohamed.

But what would be the logic behind increasing the annealing temp.? The products seem to be specific.

It will depend what are you looking for..... if you need only one band I think with a MgCl titration and an increasing of the annealing temp you could obtain one band...

Hi Christian

it is quite common to see these "ghost" bands following PCR reactions. The ghost bands are dimers of identical DNAs caused by stacking of two helices on top of each other. In your case, this is corroborated by the fact that you only see one signal in the melting curve analysis (I don't think there is any chance that a 100 bp and 200 bp DNA would show the exact same melting curve). In my experience the formation of such bands are highly dependent on magnesium and the concentration of DNA.

Anyway, nice troubleshooting by actually going the extra length to clone the DNA. By the way, when you purified the ghost bands for cloning, did you try and run them on a gel? If I am correct, you should see that at least some of the DNA "reverts" to the correct size.

Good luck

Jesper

Have you tried doing a gradient of annealing temperature and doing first 5 cycles at 61 and then rest of cycles in 55 oC, I had this problem and it solved by doing this , also denaturation step is important . I don't think that could be cause of ghost band. Also play a bit with polymerase, I tried Pfu and Taq both and for me Taq was better strangely . Good luck

If you use Phusion DNA polymerase, you need to have a shorter extension time as it polymersizes at a rate of 1 kb per 15 seconds. Also, I would decrease the cycle number to 25-30 cycles. in case rare hybrids or dimers are amplified less efficiently, by decreasing the cycle numbers you can reduce any visible amount of the undesired amplification.

This may be single-stranded amplicon, which may easily be produced if one of your primers binds more efficiently, causing enrichment of one strand, or if your PCR has run short of reagents. The decreased mobility would presumably be due to formation of secondary structures. Not my idea, I'm afraid. Another researcher suggested it in a similar discussion on Research Gate.

I found these "secondary structure" ghost bands as well but was able to get rid of them by warming my sample up before running the gel.

Hello Chritian,

I think this is a primer diamer. Please run this product on 3% agarose gel till end . It may disappear. If it appears you can do following experimentation

-Do gradient PCR using different annealing temperature.

-Use best Taq polymerase then presently use Tag

-Use Triton x 100 or BSA at desire concentration

-Run PCR at different cycles i.e.25,30, 35

-Use different concentration of Mgcl2

I hope THIS WILL PROBABLY SOLVE YOUR PROBLEM

ALL THE BEST

Dear people, I thank everyone for their time and feedback.

However, as these products are about 300 bp, there is no way that these are primer-dimer formations. I believe the suggestions of it being dimeric configurations of the target product are indeed very likely.

I currently use a commercial SYBRGREEN PCR mix, so I would be able to only increase the MgCl. But I may try the reaction with a different Taq polymerase and without a premade mix soon. I am currently not spending much time experimenting on this.

This may sound stupid but I have had artefacts like this and finally discovered that my comb was not tight enough in gel former and thus lay at an angle so that the actual band was running squint and appeared as two bands, Can't tell from your gel if this is the case as marker not seperated enough to see if they are also shadowed. But for anyone else this can happen. Comb discarded and gels now fine.

Gillian,

Doesn't sound stupid at all. Our combs and gel formers are quite old and not really perfectly straight either, and I have actually thought about this before. But like you say, it would have to appear on the marker as well and I haven't really noticed that, but you make a good point in that it's not really seperated well enough to tell in this gel. Also the combs would have to be really angled in order to see the effect which is visible in my gel, I suppose I would notice it if the combs were angled that much.

But I think it's a very usefull answer, tnx

Hello, You can find the piste of recombinated sequences. if you have sequenced yours bands separately... you can try softwares like RDP4 or DataMonkey...

 Hi Christian,

I guess that one year later we are having the same problem as you. A double band that is doble-size, although it is sometimes appearing for environmental samples and not for standards, or viceversa. Do you think it is possible these artefacts are formed only for ones, and not for others?

Jesus, we are getting crazy with these qPCRs cause meltings are perfect and we are investigating like 6 genes! Imaging how crazy could be!

Try this to avoid the problem. Increase your annealing temperature and run the gel slowly in low current 50 V. Hope this would be helpful. Share if it is so. thanks. Regards, IMJ  

These are single-stranded products that from secondary structures.  A few years ago, I excised and sequenced such bands and they had exactly the same sequence AND LENGTH as the primary product.

I would imagine it is secondary structures, I have seen then previously on my gels.  I don't get them now but a wise old technician taught me that you should allows give you samples a 1min blast on a 90oC heat block before loading them if you want nice crisp bands 

that should be always not allows 

Even some canonical double stranded DNA can form unusual DNA structures that influence gel mobility. This is particularly significant when the sequence contains longer (4-6) adenine tracts or is extremely GC rich. Normally these DNA conformations are lost when running gels at elevated temperatures (50C) which is possible in polyacrylamide but not in agarose matrix. 

Second, you can verify that the longer products are double stranded and not single stranded DNAs by cutting the PCR products with restriction enzymes. The single stranded DNAs should not be cut.

I have seen these before. These could be secondary structures of the same product not non specific bands.

I have often experienced such event. In my opinion, those ghost bands are mere electrophoresis artifacts, may be due to a delayed entry of a part of DNA fragments into the gel from the wells. I am not sure they should be due to secondary structure as they are found in all samples and at the same position in the gel.

Nicola, then how do you explain the fact that the  fragments get hung up in the well for a precise length of time and are then all released simultaneously?  They are found in all samples because all samples are the same PCR product. 

The effect of inaccurate sample loading and/or well deformation would likely result in a smear pattern rather than distinct bands.

Mark and Ales, It was just a temptative explanation: I've seen that distance between ghost band and real band is the same, independently by the amplicon lenght. I was just thinking that protein (taq) may partially block the entry of DNA in the gel. The purification of PCR product usually eliminates the phenomenon, but, actually, it is not sure for what reason.

Thanks for the comment, Nicola. Yes, you are right. DNA binding proteins certainly influence DNA mobility in gels. However, Taq polymerase is an unlikely candidate since it dissociates rapidly from double stranded DNA after the polymerisaiton.. Best. ALes

Thanks to Mark Farman for solving what, for me, is a longstanding mystery. You should publish this, Mark.

Dear Doctor,

We faced a similar phenomena  on polyacrilamide gel.

We postulated they are single stranded DNAs.

They pattern could be used for determination of several polymorphisms.

You could find these reports in

Journal of DNA and RNA Reseach 2017: 1(1): 51-60

and

Aerchives ofBiochemistry and Biophysics  2012: 525:  201-206.

Best egards

Laszlo Goth DSc

Hungary

I had same issue recently. Changing the NEB-OneTaq to NEB-Phusion-Taq got rid of the ghost bands.