Preparing samples for next generation sequencing

Hi, I’m Matt, one of the PhD students in the genetics lab. Whilst the relative simplicity, sensitivity, and scalability of Sanger sequencing means it has remained the gold standard method of characterising genetic variation, most genetics labs are rapidly adopting next generation sequencing technologies. These allow for a greatly reduced speed and cost per base, but are really only worth doing if you have a large region of the genome that you want to look at, or a lot of samples to do at once. In order to obtain high quality and reliable data from next-generation sequencing platforms, care must be taken to prepare the sample stringently.

This process of getting the samples ready is known as library preparation, where the term “library” is referring to the final set of pooled samples ready to be put on to the next generation sequencing machine. There are several different ways to achieve high-quality libraries for sequencing. Some of the stages are common to all methodologies, so we’ll focus on those here. Here we have a long piece of genomic DNA represented by this long, blue piece of modelling clay. The extracted genomic DNA will typically be more than 50,000 base pairs long. And the first step in preparing a library is to break up this genomic DNA into small fragments, because next generation sequences can usually only read short sections of DNA.

The specific size required will vary depending on the method and the application, but is usually between 150 and 250 bases. There are two main ways that this fragmentation can be achieved. The first is sonication. This involves blasting the genomic DNA with high-frequency sound waves in a sonication device. While there will be some variation between individual samples in different sonication devices, it generally takes around 30 minutes to an hour to break genomic DNA into pieces that are about 200 base pairs long. The other method uses enzymes known as transposases. These enzymes are usually involved in adding chunks of DNA into normal sequences, but a modification of their machinery means that they can cut genomic DNA into small fragments.

This takes approximately 10 minutes. So now our genomic DNA has been broken into small fragments, the next stage is the addition of adapter molecules. These small sections of synthetic DNA, represented here by our yellow and pink modelling clay, have specific sequences at their ends which adapt to the genomic DNA fragments, allowing them to be attached to the surface of the next generation sequencer. We attach these adopters to our fragments of DNA in a reaction that uses an enzyme known as a ligase. This sample is now ready for whole genome sequencing. However, in a lot of cases, this is a bit like using a sledgehammer to crack a walnut.

We might have a distinct clinical question, and therefore only a subset of genes that we want to sequence, as in the case of our monogenic diabetes patients. Or indeed we might only want to sequence the 1% of the genome that actually contains genes, known as the exom. There are a few methods that can be used in order to take this genomic sample as it is now and select a small region of it. This process is known as target enrichment, and one of the most popular methods is called in-solution hybridisation. We can think of in-solution hybridisation as being a bit like going fishing.

We use small baits, shown here in light blue, which are complimentary to small sections of our target region. And we put these in a buffer with our adapted DNA fragments at 65 degrees Celsius, and leave them to hybridise or attach overnight. The baits have small molecules attached to the back of them, shown here in white. These allow us to capture them onto magnetic beads which have complimentary molecules on their surface. We then use magnets to pull the beads out of the solution and immobilize them, meaning we can remove the regions of the genome that we are not interested in. Next, we remove the DNA fragments from the RNA baits and amplify them using polymerase chain reaction.

We have now selected a small region of the genome that is relevant to the clinical question for next generation sequencing. This whole process takes about three to five days depending on how many samples you have. This might sound like quite a long time, but if you were to sequence the same region of the genome by Sanger sequencing, it would probably take months, if not years. The next section will explain how next generation sequences work.