BSP PCR primer design explained - (Feb/11/2013 )

Hello all,

I am wondering if I could get some troubleshooting feedback.

I have been performing bisulfite conversion using the ZymoResearch Methylation Gold kit. However, after PCR, I am not getting any bands using my BSP primers. I put .3 ug of DNA into my bisulfite conversion reaction. Some other feedback I have gotten is that my amplicon is too large for BSP. My current target is 762 bps and I know that most of the literature recommends between 200-300 bps. However, I have seen papers with larger amplicons around the size of mine. Does anyone have any experience with bisulfite treatment and PCR of larger amplicons? Are there any other troubleshooting recommendations for my protocol?

Please let me know if additional information is needed.

Thank you.

Yes, the longer the amplicon, the harder the PCR. You should redesign primers to amplify a region <400 bp.

The amount of DNA you used is a bit too high, something <= 1 ug is better. Too much DNA may give you incomplete conversion.

what is your PCR conditions such as annealing temperature and cycling number?

The EXTENSION temperature should be lower as well. Many high AT regions will not extend at 72. Try extending (for longer times (double)) at 63 or 65. You are using a Taq polymerase, correct? High fidelity polymerases will not read the uracils in bisulfite modified DNA. How are your primers designed? Are you certain they are correct (it is very easy to get confused!)

Thank you very much for your quick responses!

I will definitely look at reducing the amount of input DNA.

My PCR conditions:

Step 1: 5 minutes, 94 C
Step 2: 45 seconds, 94 C
45 seconds, Tm 50 C
1.5 minutes, 72 C
Step 3: 10 minutes, 72 C
4 C, end

I have 40 cycles for step 2.

It is interesting that the extension times should be lower. I will consider that as well.

I am using a hot start Taq provided by Zymo (it is what they recommend to use with their kit).

I designed my primers manually because of some problems with the regions I am amplifying (it is a specific L1 retrotransposon in a highly repetitive region). Basically the primers that programs such as MethPrimer were giving me were hybridizing in different places all over the genome. I'll add my sequence and primers below, if you wouldn't mind taking a look at it. This is our first experience with methylation studies in our lab, so I've been trying to piece together as much advice as I can get.


gtaggaccctctgagccaggtgtgggatatagtctcgtggtgcgccgttttttaagccggtctgaaaagcgcaatattcgggtgggagtgacccgattttccaggtgcgtccgtcacccctttctttgactcggaaagggaactccctgaccccttgcacttgccgagtgaggcagtgcctcaccctgcttcagctcatgaactgtgcactgctcccactgtcctgcatctgctgtctggcactccctagtgagatgaacccggtacctcagatggaaaagcagaaatcacccgtcttctgcgtctctcacgctgggagctgtagactggagctgttcctattcagccatcttggctcctcctccagtgtagtttttatttgaagatatttccttttccacaacaggcctcaaagctctccaaatatccacttgctggttctgcaaaaagagtgtttcataactgctcaataaaaataaaggttcaactctgtgtgatgaatgcactcatcacaaacaagtttctcagaatgcttccacgtagttttaatgtgaaggtatttccttttccacaataggccccacactctgcaaatatccactaggtcattcttcaaaaaggatgtttcaaaactgctcaataaaaaaaaaggttcaactccgtgtgaggaatgtattcatcacaaagaagtttctctgaatgcttctgtgtagtttttatatgaagatatttccttttacacaatagggcacaaagggctcc

In bold: L1 promoter sequence
Underlined: Primer sequences
Rest of sequence is unique genomic sequence 5’ adjacent to L1 promoter

Amplicon size: 761 bps

L1 Primer
GTA GGA TTT TTT GAG TTA GGT GTG (TM: 51.9)

Genomic Primer
GGA GTT TTT TGT GTT TTA TTG TGT (TM: 50.9)

Again, I appreciate very much all of your input.

I don't believe those primers are correct. Let's look at a simple example. Here's a short dsDNA fragment:

5' gctagcgctagatc 3'
3' cgatcgcgatctag 5'

Complete conversion yields this:

5' gutagugutagatu 3'
3' ugatugcgatutag 5'

Note (important!) these strands are no longer complementary. You have to decide which of these two strands you want to amplify.
Let's assume we want to amplify the top strand. We'll use very short 4 bp primers (though of course you should not).
To amplify the top strand,

5' gutagugutagatu 3'

we will choose a forward primer of

5' gtta 3' (which binds the first four bases, but only starting on the second cycle of PCR, since we are amplifying ssDNA).

For the reverse primer (which will act in the first cycle of PCR) we need the primer

5' aatc 3' which will bind as the reverse complement of the chosen strand.

====
Things to notice: the forward primer has no c's. The reverse primer has no g's.
====

Now, let's amplify the bottom strand,

3' ugatugcgatutag 5'

or, rewritten,

5' gatutagcgutagu 3'

Forward primer: 5' gatt 3'
Reverse primer: 5' acta 3'

====
Things to notice: Forward primer (which binds to the END of your original sequence) has no c's; reverse primer (binding to the start of your original sequence) has no g's
These primers are different from the ones amplifying the top strand
====

You don't (in general) know which of the c's have been methylated. Ideally you should (if amplifying the top strand) choose a region of the DNA for your primers which has no c's in the primer region, both forward and reverse. This will lead to forward primers being g rich, and reverse primers being c rich. If amplifying the bottom strand, you should choose a region having no g's in the binding region of the two primers.

If you can't find a region with small numbers of c's or g's (as appropriate) then you should make sure the 3' ends of each of the primers has none of the questionably converted bases. Your L1 primer, for example, is likely pretty good as a forward primer for the top strand. The reverse primer, however, is (I think) wrong. But all of this is quite confusing, and I may have gotten it wrong.


Edit: I fixed the primer for the reverse bottom strand.

I agree with phage434. your reverse primer is not correct. it should be:

DNA: 5'- acacaatagggcacaaagggctcc
after mod: atataatagggtataaagggtttt
primer: tatattatcccatatttcccaaaa 5'

Wow, thank you, this has been incredibly helpful. I think I understand where I went wrong during primer design. Instead of taking the reverse complement of the sequence, I would only take the complement? Does this then have something to do with the strands not being complementary anymore after bisulfite treatment?

No, you never take the complement. It is always the reverse complement. The question is, what is the DNA sequence you are amplifying. The sequence is different for the top versus the bottom strand. If you concentrate only on the top strand, you can find primers this way:
1) take the original sequence
2) replace all c's with u's
3) design primers for that sequence (replacing u's with t for the forward primer, and matching a's to complementary u's in the reverse primer.

If you want to amplify the bottom strand, take the original sequence, reverse complement it. Then do the same operations as above. You will get a different result, since the c's will now be g's, and vice-versa.

phage434 on Tue Feb 12 00:50:02 2013 said: