Massively Parallel – the Next Generation of Sequencing

Sounds like a good title for a sci-fi novel, right?  What is the big deal about Next Generation Sequencing (NGS)?  Otherwise known as massively parallel sequencing or high throughput sequencing, NGS has become a technique used by many molecular labs to interrogate multiple areas of the genome in a short amount of time with high confidence in the results.  Throughout the next few blogs, we’ll discuss why NGS has become the next big thing in the world of molecular.  We’ll go through the steps of setting up the specimens to prepare them to be sequenced (library preparation), what types of platforms are available and what technologies they use to sequence.  Lastly, we’ll go through some of the challenges with this type of technology.

Let’s start with a review of dideoxy sequencing, otherwise known as Sanger sequencing, which has been the gold standard since its inception in 1977.  A typical setup in our lab for this assay begins with a standard PCR to amplify a region of the genome that we are interested in, say PIK3CA exon 21, specifically amino acid 1047, a histidine (CAT).  The setup would include primers complementary to an area around exon 21, a 10x buffer, MgCl2, a deoxynucleotide mix (dNTP’s), and Taq polymerase.  After amplification, the resulting products would be purified with exonuclease and shrimp alkaline phosphatase (SAP).  Next, another PCR would be set up using the purified products as the sample and using a similar mix as in the original amp, but with the addition of a low concentration of fluorescently labeled dideoxynucleotides.  These bases have no -OH group, so when they are incorporated into the product, amplification ceases on that strand.  Because they are present in a lower concentration, the incorporation of these is random and will occur at each base in the strand eventually.  The resulting products are then run and analyzed on a capillary electrophoresis instrument that will detect the fluorescent label on the dideoxynucleotides at the end of each fragment.  Shown below is an example of the output of the data:

NGS1

The bases will be shown as peaks as they are read across the laser.  The base in question in the middle of the picture is, in a “normal” sequence, an adenine (A), as seen in green.  In this case, there is also a thymine (T) detected at that same location, as seen in red.  This indicates that some of the DNA in this tumor sample has mutated from an A to a T at this location.  This causes a change from a histidine amino acid to a leucine (p.His1047Leu) and is a common mutation in colorectal cancers.

So all of this looks great, right?  Why do we need to have another method since we have been using this one for so long and it works so well?  There are a few reasons:

  1. The sensitivity of dideoxy sequencing is only about 20%.  This means lower level mutations could be missed.  The sensitivity of NGS can get down to 5% or even lower in some instances.
  2. The above picture shows the sequencing in the forward direction as well as the reverse direction.  This area then has 2x coverage – we can see the mutation in both reads.  If we could get a higher coverage of this area and be able to sequence it multiple times and see that data, we could feel more confident that this mutation is real.  In our lab, we require each area has 500x coverage so that we feel sure that we have not missed anything.  The picture below displays the same sequenced area as in the dideoxy sequencing above.  This a typical readout from an NGS assay and, as you can see, this base has a total of 4192 reads, so it has been sequenced over four thousand times.  In 1195 of those reads, a T was detected, not an A.  We can feel very confident in these results due to how many times the area was covered.
  3. The steps above detailed only amplifying this one area, but with colorectal cancer specimens, we want to know the status of the KRAS, BRAF, NRAS, and HRAS genes as well as other exons in PIK3CA  Using the dideoxy sequencing method is a lot of time and effort.  NGS can cover these areas in these five genes as well as multiple other areas (our assay looks at 207 areas total) all in the same workflow

NGS2

Join me for the next installment to discover the first steps in NGS workflow!

 

rapp_small

-Sharleen Rapp, BS, MB (ASCP)CM is a Molecular Diagnostics Coordinator in the Molecular Diagnostics Laboratory at Nebraska Medicine. 

Food and Drug Administration and Next Generation Sequencing

As readers of this blog are probably aware, The Food and Drug Administration (FDA) is currently considering how to tailor its oversight of Next Generation Sequencing (NGS), methodologies that can produce extremely high quantities of genetic sequences. In turn, these sequences can be used to identify thousands of genetic variants carried by a particular patient. NGS will usher in an age of truly personalized medicine in terms of patient risk assessment, diagnostics, and personal treatment plans.

Currently, the FDA approves all in vitro diagnostic (IVD) tests with the exception of laboratory defined tests (LDTs). These tests are used in clinical laboratories and typically detect one substance or analyte in a patient sample, and this result is used to diagnose a limited number of conditions. (One example would be a cholesterol test; every manufacturer that makes the analyzer and reagents to detect cholesterol in a blood samples has to get their methodology approved.) However, NGSs have the potential to detect billions of base pairs in the human genome, and therefore the potential exists to diagnose or discover thousands of diseases and risk factors for disease. Also, many NGS tests are developed by individual laboratories, not big companies, and so would be considered an LDT.

The FDA has opened a public docket to invite comments on this topic. American Society for Clinical Pathology, as well as other professional societies—American Association of Clinical Chemistry and Association for Molecular Pathology among them—has publically commented on the FDA preliminary discussion paper “Optimizing FDA’s Regulatory Oversight of Next Generation Sequencing Diagnostic Tests.” In its comments to the draft paper, ASCP stated that the “CLIA framework offers a more logical model for providing federal regulatory oversight of LDTs.” Similar points were made by AACC and AMP. The associations also agree that any regulations should not interfere with the practice of medicine.

What do you think? How involved should the FDA be in genomic testing in the clinical setting?

Further reading:

AMP comments

AACC comments

Next Generation Sequencing and Personalized Genomic Patient Care

I’m sure that everyone has heard about next generation sequencing (NGS). But why exactly is it a big deal? Even though I have spent a significant amount of time at the bench performing “wet lab” basic science research and was acquainted with the term, I did not have practical hands-on experience with NGS prior to residency. It was not a readily accessible technology during my biomedical research days prior to medical school and so I did not entirely grasp the full power of this then disruptive technology until I was a pathology resident, and even more so, as an applicant for molecular genetic pathology (MGP) fellowships.

All of us should have previously learned about the “gold standard,” Sanger sequencing. This method combined irreversible dideoxynucleotide chain termination with a detection method such as gel electrophoresis, or on a larger, automated level, capillary electrophoresis. It was powerful because it allowed us to read the genomic map that directs cellular life, albeit only one sequence at a time. It served as the mainstay for more than a quarter of a century and still is employed for smaller scale sequencing or for long (>500 bp) stretches of DNA.

In the 1990’s, initial methods of massively parallel signature sequencing (MPSS) and pyrosequencing began to appear which would lay the groundwork for today’s massively parallel sequencing (MPS), also known as second generation sequencing, or more popularly as NGS. The two most commonly utilized NGS platforms to date are based on semi-conductor technology for the Ion Torrent (now Life Technologies) and reversible dye-terminator, sequencing based synthesis (SBS) technology for the Illumina platforms.

The power of NGS comes from its ability to simultaneously sequence 1 million to 43 billion short reads (400-500 bp each). The Human Genome Project took over a decade to complete and cost nearly $3 billion whereas NGS can sequence the same genome for a cost on the order of $1000’s, a cost that is further decreasing as the technology is refined. When I was working in research (eons ago), we had nitrocellulose based dot arrays where each “dot” represented multiple copies of a specific cDNA sequence and which helped us to build expression profiles for our particular area of study. This would be analogous to analog technology and NGS would now be considered a digital version.

With conventional PCR, we could only amplify one target sequence per sample reaction. The results were also only qualitative. The results only measurable after multiple cycles of denaturation, annealing, and extension as amplified product of the expected size or no amplified product. Then came along real-time or quantitative PCR (qPCR) which some people refer to also as RT-PCR (although this is a confusing term for people like me who were around before qPCR and think of RT-PCR as meaning reverse transcription PCR). The power of qPCR was that within the linear range of detection, we could now quantitate the amount of product present in real time. NGS also provides the resolution of qPCR in terms of quantitation.

So, as a technology, NGS nicely combines and refines some (but not all) characteristics of multiple technologies with large scale profiling. But for a non-molecular person, what is the relevance? Obviously, it has taken us some time to get to this point, even though the Human Genome Project was completed in 2003 (it started in 1990). We needed time for biomedical and translational research to provide us with clinical significance to the mutations and genetic aberrations NGS could identify. We also needed to develop tools to distinguish true mutations with clinical significance from benign polymorphisms present in our population and to build our bioinformatics support.

But here we are, stepping into the new frontier of personalized genomic medicine. Sure, there is a lot of hype surrounding it and these promises will take time to keep. But these are exciting times for someone like me who fell in love with the molecular dance that plays out within our cells. One of the reasons that I want to complete an MGP fellowship is to get more hands-on practical knowledge of the nuances of NGS from a technical standpoint but also to collaborate with other physicians in directing patients to the correct clinical trials and targeted treatment – and therein, is where the power of NGS really lies. For someone who may never get to meet the patient, at least for me, there’s probably no greater satisfaction than knowing you had a pivotal part in helping a patient more effectively combat their disease. Personalized genomic medicine is another step in that direction.

[12/12/2014: edited to fix a few inaccuracies. We apologize for the error.]

Chung

-Betty Chung, DO, MPH, MA is a third year resident physician at Rutgers – Robert Wood Johnson University Hospital in New Brunswick, NJ.