I've run a couple hundred samples using a microsatellite panel, and recently some of the runs show very exaggerated stutter peaks smaller than the true allele size. To troubleshoot the issue, I've rerun the same PCR product multiple times on our ABI 3500. The problem appears to arise from the fragment analysis, rather than the PCR step, since the same product produces very different results from run to run. Some samples look fine in one run but the same PCR product shows the mystery peaks on a second run. Since some of them are showing up in each run, and the peaks are often much larger than the true allele peak (sometimes the true allele is even absent), this makes accurate scoring impossible.
I've posted three examples from three different samples and 3 different loci. The lower run for each is the one depicting the correct alleles, the upper run shows the large stutter peaks and small true peaks. In one case, the ghost peaks are 4bp smaller for a 4bp repeat, but in the other two they are 10.5bp smaller and 7bp smaller for a 3bp and 4bp repeat, respectively.
I've never had this problem before, but also am relatively new to running the ABI machine myself. I'm loading:
7ul of MClab orange size standard in SuperDi
1ul each of 3 PCR products (topped with mineral oil). They are diluted 2x first before loading.
the product was stored for ~1wk in the fridge before loading.
I've tried heating the plate to 95deg for 5 min then chilling and loading, to no avail. I've also tried re-running PCRs and the problem returns. Some plates look fine, and others are all bad.
Any suggestions on what is going on or how to fix it?
Thanks!
Kevin
SOLVED: I believe i've figured this out. I tried again with fresh PCR product, my old size standard and formamide, and a different batch of size standard and formamide. Fresh formamide + size standard fixed the problem. The old size standard/formamide mix also worked if I denatured the plate at 95 degrees C for 5 minutes followed by snap cooling at 4 deg c for 5 min.
I found this issue described here as SHADOW PEAKS, for other's future reference, on slide 58:
http://afdaa.org/2013/wp-content/uploads/2013/01/CE-trouble-shooting-Updated-AOrbison.ppt1_.pdf
Basically, it looks like my DNA was hybridizing, probably because the old formamide was bad and couldn't sufficiently denature the fragments. Looks like I just rediscovered a common problem, but hopefully this'll help someone else more quickly find the solution.
I have several hypotheses
1) wrong calibration of size scanner. Sometimes, Automatic callibration in Gene marker software (I suppose, that you use this software, as I see pictures) indentify incorrect peaks as size scanner. This problem usually osccurs, when some microsatellite peaks interfer with size scanner. You have to perform callibration manually. I usually perfmorm maual correction of size scanner in one third of samples.
2) interaction between sveral loci in multiplex - sometimes, locus works good, if it is alone, but irregular with other loci
3) signal from peaks from different color - you have loci marked by several fluorescent cololors. These colors have largest signal in one color, but also secondary (incorrect) singals in other colors. This problem is especially important, when markers strongly differs in quantity of pcr product - false signal of one locus can be larger than good signal from other alelle. Therefore, check what you will see if you vizualize other colors.
Hi Michael,
Thanks for your suggestions. Unfortunately, I think i can rule them all out.
1. I thought of this, and have this problem occasionally, but the shifts only occur in some loci, not others, for each individual fragment analysis sample pool. Plus, the shifts are of different number of bp for different loci. If it were a size standard problem, I think it would affect all loci the same.
2. Do you mean multiplex in PCR or in fragment analysis? PCR multiplex isn't the issue since it's not a PCR issue, and i have this problem with both multiplex and singleplex PCR reaction products. But i hadn't considered that they could interact with each other in the fragment analysis run. I'll do a trial to see if running one locus separately from the others clears up the issue.
3. It's not pull up from other dyes, i checked for this.
Thank you for thinking about this, i really appreciate you taking the time.
Microsatellite markers, also called short tandem repeat (STR) markers, are polymorphic DNA loci that contain a repeated nucleotide sequence. Each repeat unit can be 2 to 7 nucleotides in length, and alleles differ by the number of repeats. The number of nucleotides per repeat unit is the same for a majority of repeats within a microsatellite locus.
A PCR primer pair consists of two oligonucleotides (forward and reverse primers), typically 15 to 30 nucleotides long. Each primer hybridizes to its respective complementary strand of the DNA template such that the primer pair flanks the region of interest. Based on the application, one or both of the primers may be labeled with a fluorescent dye.
Hi Kevin,
Did you load PCR products into new or used ABI-plates? We observed cross-sample contamination when re-using ABI-plates that were not properly cleaned.
Best
Hi Mariana,
Thanks for thinking about this! Yes, these primers were designed for this species. Frustratingly, the primers/mixes worked fine a year ago, so something in my reaction protocol has changed (i suspect with the ABI machine) between then and now.
Best,
Kevin
HI Gwenael,
Thanks for this suggestion. The ABI plates were cleaned and re-used. I'll try a clean new plate next time i can go into the lab, but since the problem isn't one of cross contamination (the stutter peaks correspond to the correct sample genotypes closely, not to the genotypes of different individuals), I'm not sure if it will help. But it's worth a try!
Thanks,
Kevin
- Badges
- Science method