Product Survey: NGS library preparation kits

Proper Preparation Prevents Poor Perfomance
by Harald Zähringer, Labtimes 04/2015

Garbage in, garbage out − that simple wisdom holds especially true for next generation sequencing. Commercial library prep kits promise to create complex, representative, non-biased NGS sequencing libraries. There is, however, still plenty of room for improvement.

The construction of libraries for next-generation sequencing (NGS) is comprised of a few simple steps (at least in theory). Since all of the necessary enzymes and components are well known and easily available, NGS library preparation looks like the perfect playground for “home-brewers” and tinkerers. However, “kit-addiction” amongst life science researchers has also taken its toll in NGS library-making and many (if not most) researchers are hooked on commercial kits.

Traditional NGS library construction, based on DNA as starting material, requires the following steps: end-repair of physically- or enzymatically-prepared DNA fragments, phosphorylation of 5’-ends, A-tailing of 3’-ends, adapter ligation and, last but not least, a PCR to enrich for fragments with correct adapters on both ends.

Nothing fancy

Hence, commercial kits usually contain T4 polynucleotide kinase, T4 DNA polymerase and the large fragment of the Klenow enzyme for fragment end blunting and 5’-phosphorylation. They also include Taq DNA polymerase or Klenow Fragment (3’-5’ exo-) for A-tailing of 3’-ends (i.e., adding an adenine) as well as T4 DNA ligase, which joins the tailor-made adapters (supporting the favoured NGS platform) to the modified 5’ and 3’-fragment ends. Finally, a thermostable DNA polymerase is enclosed in most kits for PCR amplification of NGS-ready fragments via adapter-specific primers. That’s it! No fancy enzymes or weird buffers, just plain vanilla molecular biology stuff.

Fragmentation, followed by adapter ligation, is still the standard technique for library construction. Increasingly, however, it is being replaced by an alternative transposase-based method, enabling DNA fragmentation and adapter attachment in a single ‘tagmentation’ reaction.

PCR-bias during NGS library preparation may completely ruin sequencing results. Researchers try to circumvent bias with PCR-free library preparation kits.

Fast but expensive

Tagmentation kits utilise a hyperactive Tn5-transposase derivative (Nextera), which fragments the DNA and simultaneously attaches the adapters to both ends. Tagmentation considerably accelerates NGS library preparation but the costs per sample are still in the range of €50 − which is actually more than three times the price of sequencing per sample.

You may, however, cut the costs considerably, by simply ignoring some of the guidelines given in the manual of the Nextera kit. A team centred around Roy Kishony from the Technion-Israel Institute of Technology in Haifa, and Michael Desai from the Harvard University in Cambridge, USA, established a simple protocol for inexpensive, multiplexed, library construction based on a Nextera kit (Baym et al., PLoS ONE 10 (5): eo128036).

The group divides their protocol into five modules: standardisation, tagmentation, adapter and barcode PCR, PCR clean-up and, finally, quality control and pooling.

Reduced volume

Let’s focus on the second and third modules, which are performed applying a commercial Nextera kit. Reducing the costs in the tagmentation module is a complete no-brainer. All you have to do is to lower the reaction volume from 20 µl in the standard Nextera protocol to 2.5 µl (you may find detailed step-by-step instructions for every module in the paper). Obviously, the smaller reaction volume has no adverse effect on the diversity of the sequenced DNA for megabase-sized genomes; however, a slightly raised volume may be necessary for larger genomes.

Honing module three to lower costs and enhance efficiency is also no big deal. During this step, the tagmentated DNA is amplified with limited PCR cycles. The PCR also allows inserting adapters required, e.g., for binding to the sequencing platform or barcoded adapters necessary for multiplexed sequencing approaches.

Fine tuning the protocol
Baym et al. use either standard or custom-made, barcoded adapters in this step, however, they substituted the PCR reagents provided in the kit with third party library amplification reagents. “In principle,” as stated by Baym et al., “Any hot start, high-fidelity enzyme with low GC-bias amplification should work.”

The group also modified the PCR programme given in the Nextera manual and recommends a longer initial denaturation and shorter extension times as well as more PCR cycles to adapt the PCR to the altered conditions.

The optimised Nextera tagmentation protocol reduces the library preparation costs per sample to approximately €8. However, the risk of PCR amplification bias in GC-rich genomes is not eliminated. The easiest way to completely avoid PCR bias in critical probes is by using PCR-free kits, which are offered by several vendors.

Kicking out the PCR

PCR-free library preparation kits are based on adapters containing all necessary elements for binding to the NGS platform. The adapters are ligated to the DNA fragments with high efficiency, eliminating the need for a further amplification step. However, PCR-free kits usually require high amounts of input DNA (500 ng to 1 µg), disqualifying them for library preparation from cell-free DNA (cfDNA).

A Swedish team, lead by Magnus Nordenskjöld and Sten Linnarsson from the Karolinska Institut in Stockholm, recently fixed this problem (Karlsson et al., Genomics 105, 150-8). The nordic PCR-free amplification strategy utilises biotinylated-linkers and streptavidin-coated beads to ligate the adapters to cfDNA molecules.

Immobilised ssDNA

In the first step of the protocol, double stranded cfDNA is denatured and dephosphorylated. A biotinylated linker is ligated to the 3’-end of the ssDNA; the ssDNA-linker complex sticks to streptavidin and is immobilised on the streptavidin beads. Subsequently, the first adapter is added. It is composed of three parts: a platform-specific sequence at the 5’-end, a short stretch complementary to the linker-DNA in the middle and a random nucleotide overhang at the 3’-end. The adapter hybridises to the linker and is extended by a polymerase.

Low DNA input

The second adapter, introduced in the next step, is double-stranded with a 5’-phosphate at the lower strand and a 3’-dideoxy nucleotide at the upper strand. After ligation of the second adapter to the double-stranded DNA, the strands are dissociated and released from the streptavidin beads for subsequent sequencing.

According to Karlsson et al., the Swedish PCR-free procedure works with just 50 ng of input DNA. So, it may be worth a try − and fiddling around with the protocol to optimise it further is most probably more fun then just buying another NGS library preparation kit.

First published in Labtimes 04/2015. We give no guarantee and assume no liability for article and PDF-download.

Table of Products as PDF-download: pdficon Formatted for Printout