How to prepare samples for multiplexed amplicon sequencing on Illumina systems?

There are multiple valid protocols available for amplicon sequencing on Illumina systems.  Here we describe one of many options: A  two-step PCR protocol to generate complete sequencing libraries.
This protocol has the advantage that it does not require custom sequencing primers and that the barcode-indexing oligos can be re-used for multiple different amplicons and future projects.   We suggest to follow a  “16S amplicon” protocol that was explicitly designed by Illumina to be adaptable to other targets (please see the full protocol and pages 3 and 4 here).

Once you have designed the oligos as described in the Illumina protocol (forward overhang plus your sequence-specific primer as well as reverse overhang plus sequence-specific primer), we suggest checking these sequences on the IDT oligoanalyzer ( https://www.idtdna.com/calc/analyzer ) for secondary structures. It is advisable to avoid any sequences that generate a Delta G smaller than -9 for any of the structures.

There is no need to purchase an Illumina Nextera index kit. The sequences for the index primers (26 i7 index 1 sequences; 18 i5 index 2 sequences) are available on pages 7 and 8 here. These indices allow for the combinatorial sequencing of up to 468 samples. When ordering oligos please use the index sequences in the “Bases in Adapter” columns.  The oligos are used for standard PCR reactions. Thus, low-cost desalted oligos can be ordered for this purpose anywhere and will work just fine. We strongly recommend using plates with single-reaction aliquots of these index primers for your experiments to make sure that index primer stocks cannot become contaminated.

Your first round PCR amplicon products will have universal tails/tags/overhangs on both ends.    Since you can use dual indexes, you could order for example 5 index oligos with i5 indexes and 5 index oligos with i7 indexes and have 25 usable barcode combinations for your project.  If you are using single indices they have to be i7 (P7 adapter) indices.  However, for HiSeq 4000 and NovaSeq sequencing you should use uniquely-dual-indexed (UDI) barcode combinations.

The first round PCR primer designs use Nextera-style tag sequences (overhang sequences) and look like this:
Forward overhang P5-tag: 5’ TCGTCGGCAGCGTCAGATGTGTATAAGAGACAG-[locus-specific sequence]
Reverse overhang P7-tag: 5’ GTCTCGTGGGCTCGGAGATGTGTATAAGAGACAG-[locus-specific sequence]

The second round PCR primers are Nextera-style index primers –  i5 and i7 indicate the location of the barcode index sequences:
P5-PCR index primer: 5’ AATGATACGGCGACCACCGAGATCTACAC[i5]TCGTCGGCAGCGTC
P7-PCR index primer: 5’ CAAGCAGAAGACGGCATACGAGAT[i7]GTCTCGTGGGCTCGG

Please optimize the conditions of the first round PCR to avoid primer-dimer generation.  The PCR reactions should be cleaned up with Ampure XP beads (or similar) and resuspended in EB buffer.
Once you have verified (via agarose gel electrophoresis) that the PCR products for all samples are clean and of about the same and expected size, the samples should be pooled equimolarly. We suggest to quantify the samples via fluorometry (Qubit or plate reader) for accurate pooling.

In case you are targeting only a single amplicon, it helps to create sequence diversity by adding a set of PCR primers with added diversity spacer “N” bases (or defined bases; up to seven of them) between the overhangs for both forward and reverse primers (Fadrosh et al. 2014, Wu et al. 2015). The resulting set of primers should be pooled in equimolar ratios and used for the first round of PCR.

The original Illumina design looks like this: overhang+locus-spec. sequence (no spacer): 
                                     5’ TCGTCGGCAGCGTCAGATGTGTATAAGAGACAG[locus‐specific sequence]
Complementary stagged spacer versions of this oligo would be:
One spacer base added:               5’ TCGTCGGCAGCGTCAGATGTGTATAAGAGACAG‐X-[locus‐specific sequence]
Two spacer bases added:              5’ TCGTCGGCAGCGTCAGATGTGTATAAGAGACAG‐XX-[locus‐specific sequence]
Three spacer bases added:            5’ TCGTCGGCAGCGTCAGATGTGTATAAGAGACAG‐XXX-[locus‐specific sequence]
Four spacer bases added:             5’ TCGTCGGCAGCGTCAGATGTGTATAAGAGACAG‐XXXX-[locus‐specific sequence]
Five spacer bases added:             5’ TCGTCGGCAGCGTCAGATGTGTATAAGAGACAG‐XXXXX-[locus‐specific sequence]
Six spacer bases added:              5’ TCGTCGGCAGCGTCAGATGTGTATAAGAGACAG‐XXXXXX-[locus‐specific sequence]
Seven spacer bases added:            5’ TCGTCGGCAGCGTCAGATGTGTATAAGAGACAG‐XXXXXXX-[locus‐specific sequence]

Knowing the locus-specific sequence one can certainly be smarter and make the two “N”s directly before the locus‐specific sequence different from the first two bases of the locus‐specific sequence (Fadrosh et al. 2014). If pooling amplicons for multiple targets (more than 8) there is no advantage using diversity spacers.
Some downstream programs might require the removal of the diversity spacers.  dbcAmplicons can demultiplex the data as well as trim/remove the diversity spacer.

Fungal ITS: Illumina has published a second version of this protocol, modified to sequence and study fungal ITS sequences.
Qiagen offers a commercial amplicon prep kit for multiple 16S regions and ITS for which they have perfected the diversity spacer approach described above. This kit eliminates the need for PhiX spike-ins.

A much more detailed protocol for 16S and other amplicon sequencing is available here: Gohl et al. 2016

Please see this page for the library requirements for sequencing (http://dnatech.genomecenter.ucdavis.edu/sample-requirements/). The above protocol will generate a surplus of library material.


Category: 03 Sample Preparation & Sample Requirements

← FAQ
Posted in
Latest Tweets
  • Yesterday, the lab got trained on a Mantis (liquid dispenser). Today this visitor shows up and climbed up the stra… ,
  • Writing a great paper may be not enough? Producing an accompanying youtube video might help? An new scRNA-seq prot… ,
  • One of many CCGP reference genomes to become available in the next months - based on work from our long-read sequen… ,