Dear Meelad,

This all sounds like a good setup.

I'd like to make some points:

1. SNPs is a single nucleotide polymorphism, ie - you would only have a single base difference. You obviously are looking at an indel (insert/deletion) of about 50 bp difference. As you are looking at the genomic DNA and thus do not need quantification of your isoforms, I personally would not even bother to do quantitative PCR, but would simply run a normal PCR and then put the products on a gel. That should be cheaper too.

2. If you saw primer dimers on the gel, then this explains the weird melting curve and amplification mixture. Although, in my experience, the switch from specific to unspecific amplification is a bit more gradual than 1 degree - more like 2-4 degrees difference.

Perhaps you also had some contamination? Very easy to do if you are looking at DNA...

Was there a reason you ran it at 59 degrees? I'd maybe do a temperature gradient PCR and then pick the highest temp where you still get good amplification.

3. Crowded street shouldn't make a difference - you'd need some serious vibrations to throw the machine off. The electric supply sounds like the more likely problem. Could be that during the run, it fluctuated and resulted in not quite the correct (high enough) temperatures?

4. If this problem has not re-occured regularily, but only is a once a month occurence, I wouldn't spend too much time troubleshooting. Just repeat the relevant samples.

Do you have a power conditioner/surge protector attached to your machine?  Sometimes inconstant/fluctuating amperage from the power source/outlet can cause strange anomalies like this.

Dear Jack

yes the machine is connected to power conditioner but i doubt about its accuracy , please i want to know is there another reason like solution problems(master mix, primer concentration ,or DNA validity) or some mechanical reason (before a while my plate was stuck and the door couldn't be opened till the technician opened the instrument from backward but i used it twice and the results were O.K)

please i want to know  all the reason(s) for such problem

Thank you

Without knowing the kind of machine - it is hard to answer.  Are you using an Mx4000? When you say you are dealing with two primer fragments - exactly what do you mean?  Slight mismatches in primer sequences with target can result in these kind of jagged amplifications.  Use log scale on your y-axis in picture 1 and you will see that the problem seems less-pronounced.  There are a number of things that might cause this - and not all are explainable.  If your master-mix has been freeze-thawed too many times could be an issue, and SYBR Green can go bad over time.  What mastermix and machine are you using?  How do you prepare your DNA?  Are you sure that your DNA has not been degraded ... etc.

Meelad,

We need a bit more information to help with troubleshooting...

"human SNP resulting in two primer fragments" as Jack says, could you explain in a little bit more detail? Do you digest with a restriction enzyme? Or do you have a mix of primers that recognize the different isoforms? Do you detect the SNP by meltcurve?

Are you using SYBRGreen or Taqman assays?

Your amplification curves do not look too bad (except the three that come out lowest at the final cycle - looks like inhibition going on in those). Inhibition may result from contaminants in your DNA purification method (how do you get your DNA and what is the quality?)

Looks like there is quite a bit of variation in when the plateu phase is reached - do you use a Mastermix and then aliquot out for all your samples & replicates? What is the volume of your PCR reaction? And how much DNA do you add (volume and mass?)? I can't tell which of the curves are replicates of the same sample as they are all the same colour. Seeing the difference in the replicates versus difference in samples would help with troubleshooting... Sometimes, large variation can be caused by faulty pipets - when were yours last serviced? You can check accuracy of pipets by pipetting ten replicates of water onto sensitive scales and determining the CV of the values...

Now to the more obvious problem: the melt curves suggest that you have serious issues with unspecific amplification. Melting around 63 would suggest primer dimers, then you have this broad shallow peak between 70-77 degrees which is not right at all and suggests multiple fragment amplification. Do you use a Hotstart polymerase? Have you run some of these PCR reactions on a gel? What do your good results look like?

Other issues I could speculate on - is your PCR machine situated on a bench that also contains a centrifuge or pipetting robot or something that could cause vibrations? Or are you next to a road that is occasionally used by some very heavy vehicles?

Did this problem ever occur before the machine was stuck? It could be that it is now slightly misaligned (or something is loose) and whenever there is some vibration this misalignment results in readings that are only partly from the wells?

Really need some more information Meelad...

Dear Jack & Petra

·        I’m using Bioneer Exicycler 96 RT-PCR

·        What I meant by “ two fragments” that when there is a heterozygous SNP (208 bp & 259 bp), I get two fragments on gel but when I have homozygous one I  get only one fragment

·        The Mastermix I’m using GoTaq® qPCR Master Mix is a ready-to-use 2X master mix for use in real-time quantitative PCR (qPCR and RT-qPCR). The system contains BRYT Green® dye, a novel fluorescent DNA-binding dye.

·        Usually I prepare fresh mastermixes and only twice a time the Mastermix is freeze-thawing but the working solution primers and my DNA sample were freeze-thawed up to approximately 3-4 times .

·        My DNA samples were formerly run into gel electrophoresis and they were intact.

Dear Jack & Petra

·        I’m not using REs in my research but only primers to detect certain human SNPs

·        My detection depend upon melting curves and gel electrophoresis

·        My PCR Mastermix volume is (45μl )and the DNA samples (5 μl) as indicated in company instruction

·        The protocol has a hot start polymerase reaction , and I previously ran my samples (with annealing/extension temp.=60C/ 40 cycles) but actually I didn’t get primer dimers ;but this time the ann./ ext. temp. was 59 C. and I got primer dimers unfortunately.

·        The machine located on a bench with no other vibration- causing tools ; but it’s in front of crowded street.

·        Well, I doubt about the lab electricity power it’s sometimes fluctuated and the power conditioner is not good.

·        This problem occurred previously when there was a problem with the plate homogenizer .

·        To be frank with you I ran twice a time my samples after the problem of stuck plate and the results were O.K (as I mentioned the door couldn’t be opened till the technician opened the machine from the back and pulled out the plate).

Very much grateful to you if you support me with your advices

Dear Meelad,

This all sounds like a good setup.

I'd like to make some points:

1. SNPs is a single nucleotide polymorphism, ie - you would only have a single base difference. You obviously are looking at an indel (insert/deletion) of about 50 bp difference. As you are looking at the genomic DNA and thus do not need quantification of your isoforms, I personally would not even bother to do quantitative PCR, but would simply run a normal PCR and then put the products on a gel. That should be cheaper too.

2. If you saw primer dimers on the gel, then this explains the weird melting curve and amplification mixture. Although, in my experience, the switch from specific to unspecific amplification is a bit more gradual than 1 degree - more like 2-4 degrees difference.

Perhaps you also had some contamination? Very easy to do if you are looking at DNA...

Was there a reason you ran it at 59 degrees? I'd maybe do a temperature gradient PCR and then pick the highest temp where you still get good amplification.

3. Crowded street shouldn't make a difference - you'd need some serious vibrations to throw the machine off. The electric supply sounds like the more likely problem. Could be that during the run, it fluctuated and resulted in not quite the correct (high enough) temperatures?

4. If this problem has not re-occured regularily, but only is a once a month occurence, I wouldn't spend too much time troubleshooting. Just repeat the relevant samples.

Dear Petra

I do appreciate your advices , but I’ll explain for you some points:

1.     I used RT-PCR to detect certain SNPs since it’s a throughput method for monitoring the melting point and detecting certain allele(s)

2.     Optimization for temperature gradient was performed  formerly and I observed the best temperature to detect the heterozygous SNP was between 58ºC-59 ºC, but actually I’ve found recently that there is a problem concerning the instrument itself the Ct threshold increased “sharply” from 1300 to app. 5000 and there was “weird” fluorescence and melting curves levels (I’ll include attachment for these recent results).

3.     The electric fluctuation is more often to be one of the main causes of unreliable results.

4.     I decreased the vortexing and mixing duration (as I read in some manuals’ instructions) and I got lesser “jagged” curves , but the problem of not having clear results for heterozygous allele melting curves and gel band still exist.

Thank you so much for your interest and precious opinions.

On your machine, has someone else (another user of the machine) adjusted the "gain" setting on the SYBR Green filter/photomultiplier tube on the machine?  This would change the threshold setting directly.  Also helpful, would be to use log scale on your y-axis, instead of linear scale (in your first attachment above), this way you can truly see the log-linear region of your amplification plots. The region you seem to be concerned about (the far plateau region) is (in my opinion) way outside the area of concern for this assay.

Jack makes a number of good points.The plateau region is not usually of interest as you would quantify with Cq in the very early amplification phase. I'd add that perhaps you need to check the settings for your melt curve on the machine - if someone changed something in your program for example the cooling down after the melt curve, this could result in weird re-annealing that would appear as smear or indistinct bands on the gel. However, that would not explain your weird melting curves. You could try and put the PCR products onto a gel without running the melt curve first, or at least take out some product and run in parallel with samples that have undergone the melt curve.

As you are using the PCR products for detection of heterozygosity, it may be worthwhile to reduce the cycle number to 30 or 35 to avoid unspecific amplification in the late stages. In PCR, you will amplify something sooner or later... question is if it is a specific product OR amplifying primer dimers, non-specific products or contamination.

A  few questions for you: What do your water controls look like? Did you get a new lot of PCR mix or primers? Do you change your aliquots of consumables after this happens (it could be contamination - as I said very easy to do when looking at DNA? Did someone else prepare a fresh lot of aliquots (I'm wondering if your primer concentration is a bit too high)? And is it better after you change aliquots?

Dear Jack

Thanks for interest , 

To answer your question , yes there was someone else who has recently used the machine for gene expression but with also  Sybr green , well, actually you may be right because when this person started to use the machine i really noticed such "weird" curves .

Regarding the log scale , i am going to try it in next run .

Dear Petra

Thanks for interest

- For explanation , my program is intact and i check it every time i perform the run, but as i inform Jack in the former reply , there was another person who used the machine with gene expression.

- Well, i doubt about decreasing the number of cycles because i'm following the company protocol .

- I didn't catch your question about water control ?! i'm using nuclease- free water for preparing working solutions and TE buffer for the stock solutions.

- I didn't change any of my aliquots but each time i consume my working solutions , i prepare new ones from the stock primer i have .

- My stock primer solution concentrations  are= 100 uM

and i prepare the working solutions with concentrations= 10 uM ( as the company instruction).

Meelad,

a water control in PCR is when you add everything but add NO template (DNA in your case). It will tell you if you have contamination in your solutions or during preparation.

I personally do not use TE buffer in any PCR solutions. It only contains 1 mM EDTA and you dilute your stock solutions in water, so the final concentration will be very low. However, EDTA is a chelator - it will bind cations that enzymes like DNA polymerase need to work efficiently.

Regarding your stock primer - I would aliquot this so as not to thaw/freeze it too many times if you repeatedly go back to the stock solution to make working dilution. You could also make a large volume of 10 uM working solution and aliquot it in tubes @ 2-3x volume of what you usually use in one PCR.

How much of your 10 uM working primer solution do you add per 100 ul PCR volume?

Every time I have "funny" PCR results or high water control amplification, I throw away my water aliquot, my primer aliquots and my PCR mix aliquot. No use wasting time trying to figure out where contamination comes from - and I mostly work on gene expression that is very specific to certain tissues, so would not be found in normal skin cells or dust.

When working with DNA, one has to be so much more carefull as skin continually sheds dead cells and all body cells contain the same DNA...

Dear Petra

-For using TE buffer, i 've read  that the stock primer solution should be dissolved in this buffer so as to protect the oligos from degradation , but the working solution was diluted with dH2O in order to lessen the effect of the chelating substance(as you mentioned).

-Yeah , you're right about the repeated freeze/thawing of stock primers i did really miss that .. i should've  prepared larger volume for that. 

-For qPCR mix, the final volume (as i mentioned previously is 50 uL, , and i usually add  only 1 uL of the 10 uM working  solution (regarding to my procedure).

-The NTC , i did previously worked with and there was no contamination at all as the gel interpretation results , but next time i 'm going to include it in my experiment.

Thanks