For best long-range PCR results for RNA isoform profiling the RNA quality (RIN) should be as high as possible (>7), although we have found RIN >6 works for many genes. The longer the isoforms of interest and thus the amplification length, the more important high RNA quality is.
Capturing RNA isoforms - how many primers do I need?
To ensure capture of known and unannotated RNA isoforms, you may need to design multiple different forward/reverse primers to cover alternate start/end sites. UCSC-GB tracks that can support primer design and selection include:
Comprehensive Gene Annotation Set from GENCODE: All, non-coding, splice variants.
Human mRNAs from GenBank.
NCBI RefSeq genes.
FANTOM5: CAGE reads.
Human ESTs.
Using the above tracks can provide information and evidence to support design and placement of primers. For example, CAGE tags help identify bona fide transcriptional initiation sites. We suggest placing primers in the 5' and 3' UTR regions close to the ATG and STOP to minimise UTR sequence differences. If these areas are unsuitable for primers the first and last exon can also be used, however isoforms that use alternative exons will be missed. Avoid designing primers across exon-exon boundaries/junctions, as alternative use of either exon may result in amplification failure of specific RNA isoforms. Additionally primers that cross exon-exon boundaries may be able to prime and amplify solely from the 3' exon. This means you don't actually know what 5' exon was present and potentially what isoform was detected.
Note: it is a good idea to test primer pairs individually first to check amplification. E.g. individually test each forward primer with a single reverse: F1-R1 and F2-R1, prior to combing: F1 & F2 with R1.
Taq polymerase & PCR optimisation
There are several DNA polymerases available that are capable of amplifying long products. In some cases it is important to try a different polymerase if one fails to amplify the expected product. Selection of a suitable polymerase can depend on several factors including reaction speed (min/kb), cost, amplicon size and fidelity. In most cases we suggest using test RNA/cDNA first to optimise PCR conditions, considering overall cycles, annealing temperature (test gradient) and 2- or 3-step protocols depending on primer lengths.
Generally, PCR cycles should be kept to a minimum to reduce PCR bias and artifacts, however there needs to be enough PCR product post bead clean-up to ensure reliable ONT barcoding. We commonly barcode 1 ng of PCR product, so the yield required is low, helping keep PCR cycles down.