PCR primer design - published primers trustable? - (Jul/27/2012 )

Hello,
I am trying to use published articles to design primers for four receptors that I am interested in. However when I use the NCBI RefSeq (mRNA) to confirm that the primer works it often produces an error saying that the primer is not found within the sequence. I finally found one paper that gave the RefSeq (mRNA) that they used and that particular sequence was "obsolete". So here are my questions:

1) Should I not worry that I can't confirm the sequence and just trust that since the article has been published that it will work?

2) Should I even worry with published articles and just design my own based off the updated NCBI RefSeq (mRNA)?

3) Is there a way to blast primers against the entire genomic sequence? (I would have to invert the primers sequence first, correct?)

Thanks for any help!

NO, Don't trust any primers published in papers. The primer sequences could be wrong either due to bad design, sequence identification, or errors in making the paper.

Yes, I would first try to design my own primers. Published qPCR primers from high quality papers proven to work may serve as references

Yes, use UCSC in silico PCR tool. But be cautious that if a primer span exon-exon junction, the tool may not return a hit.

1) I agree with pcrman, never trust, check first. Some people have bad habits publishing primers with errors deliberately, and with majority of ordinary papers I just wonder how their primers could even work.

3) Use PrimerBLAST to check (blast) your primers with selected databases, it's a perfect tool for finding nonspecific binding of a primer pair not just oligo alone. It's primary intended for primer design, but if you fill in your primers, it will just check them. Also shows intron spaning.
No need to invert anything (you don't need to that in BLAST either, searches on both strands) just be sure you put there sequences 5'->3'.

prcman and trof thank you very much for such excellent advice.

The issue I'm having with primerblast is that the primer sequences I'm using are based off obsolete versions of certain genes.

For example. In the lab we have be using a primer for the Adora2a gene and it produces good results on mice cDNA. Here is the sequence:
A2A F: AGCCAGGGGTTACATCTGTG
A2A R: TACAGACAGCCTCGACATGTG

However, if you were to blast it against mRNA RefSeq: NM_009630.2 it doesn't work. Yet, go to UCSC In-Silico PCR (excellent reference pcrman) then you would get a result on Chromosome 10 where the Adora2a gene is located.

So here is my main problem. I can only find primers in published articles that were designed on obsolete sequences ( ex: Adora 3: xm_131085) or primers that were designed on sequences that are not easily found (Easily found: Adora2a NM_009630.2 ; Difficult to find: Y13346 (from old paper)). So how do I resolve this? Logically you would think that the gene or mRNA sequence never changes but apparently it does.

What is the purpose of the example primers for Adora2a? the primers bind to the 2nd intron of that gene and is not apparently for RT-PCR. that is why you did not get hits blasting against RefSeq.

http://genome.ucsc.edu/cgi-bin/hgTracks?hgsid=288506207&db=mm9&position=chr10:74789100-74789262&hgPcrResult=pack

Interesting! The purpose is for PCR, we've been using for awhile with good results. The guy who made them is no longer with us. How did you find that it binds to the 2nd intron?

I know it is for PCR, but for what type of PCR? Genomic PCR or RT-PCR?

jeee on Wed Aug 1 03:29:49 2012 said:

If you BLAST your forward primer with Mouse genomic and transcript database and limit search with "Adora2a" query it only get 100% hit on a genomic sequence in Adora2a gene, no binding in transcript, so it must be in intron or untranslated part.

Reference sequences may change a bit over time, but mainly the untranslated part or minor polymorphisms have some small changes. Usually that's not a major problem in primer design. And it's generaly not that important on what sequence the primers were designed, important is if that sequence corresponds with the reallity, so basically.. BLAST.

If you used this primers for expression studies on cDNA, and got a signal, I'm affraid what you got is genomic contamination. Only way to get intron signal on PCR is with DNA or unspliced RNA from nuclear fraction, not mature mRNA.

Thanks guys, these are very good points. I am in the process of going back over the data and throwing out anything that was meant to be mRNA data that was used for the primer above (primers on intron). I really do appreciate the help!