How do the Nextera Read1 and Read2 primers both work

[Phage434]

Does anyone understand how the Nextera Read1 and Read2 primers can work when they both have identical 19 bp regions at the 3' ends?

I would never think about designing a Sanger sequencing primer with that structure, nor a PCR primer. It's hard to believe these primers are possible with normal base composition.

The hinge sequence on Truseq primers has a similar issue, but with "only" 13 bp of identical sequence.

[kcchan]

The sequencing read primers aren't PCR primers. They're added separately at the beginning of each sequencing read to start the synthesis reaction (ie. Read 1 primer at the start of read 1, read 2 primer at the start of read 2). The templates themselves are also single stranded at this point and anchored to the flow cell by the P5 and P7 Illumina adapters.

The actual primers used in the PCR hybridize to a region upstream of the identical sequence to avoid problems with improper adapter orientation and dimerization.

[Phage434]

I understand those things. I still don't understand why a Read 1 promoter does not bind efficiently to the Read 2 site (and vice - versa).

[kcchan]

It probably does work, but it wouldn't make a difference in sequencing. All of the DNA clusters on the flow cell are single stranded and in same orientation as determined by the flow cell primer. Therefore it's unlikely you will get second read sequences contaminating your first read or vice versa.

[Phage434]

I see. To clarify for others, during forward read, the flow cell has only ssDNA, and the Read 1 primer can only bind to the Read 1 site. The other potential binding site is only present in a reverse complement form. After finishing the Read 1 phase, the flow cell is switched to containing the reverse complement of the original ssDNA, cleaving the original ssDNA and washing it away. The Read 2 primer can now bind. So, in principle, the Read 1 an Read 2 primers could be identical, and we would still get directional sequencing.
Thank you.

[kcchan]

Yep you got it. The adapter sequences themselves are this way by necessity rather than by design. The Nextera enzyme inserts a specific transposed sequence of DNA during the tagmentation process, which forces you to use this sequences as your primer. The way directionality is done is by having two different sets of enzymes charged with different 5' overhangs on the transposed sequence. During PCR the primers enrich only fragments with different overhangs at each end and also add the Illumina adapter primers sequences to both ends.

The story is a bit different with TruSeq adapters. Here the identical sequence is necessary to form the double stranded sequence necessary for adapter ligation. 5' and 3' overhangs are used to add the full Illumina adapter sequence at each end. This improves efficiency over having two separate adapters as all ligated DNA form sequencing compatible product on both DNA strands, so the theoretical efficiency of the TruSeq adapters is 4x that of having two separate double stranded adapters.

[Phage434]

Point taken, but I believe it is only a factor of two. With adapters A and B you would get all combinations, AA, AB, BA, BB. AB and BA both bind and amplify in the flow cell. This also explains how the Nextera random insertions work, which also yield a similar distribution. Only half of the resulting fragments will bind and form clusters.