Next-generation sequencing (NGS) systems were developed following the human genome project, and have been repurposed from their total genome sequencing, metagenomics, and plant genetics origins to multiple new and exciting life science applications; including diagnostic purposes in HIV resistance and identification of new tumor clones in oncology. As the application of genetic sequencing reaches into new branches of basic science and clinical diagnostics, laboratories will have to address the logistics of consistently running the high-throughput DNA extraction, shearing, cleanup, amplification, and sequencing required during each run. Questions of quality control, cost-effectiveness, and management of sizable data sets impede many labs from adopting automated NGS platforms. This article will outline the many advantages of automation, address disadvantages, and provide examples of current and future NGS systems and applications of high-throughput automated sequencing.
Automated DNA sequencing: Goal-driven development
The first automated DNA sequencers were produced by the Leroy Hood Laboratory at Caltech and the Wilhelm Ansurge laboratory at the European Molecular Biology Laboratory, and ultimately commercialized by Applied Biosystems, Inc. (Foster City, CA) and General Electric Healthcare (Waukesha, WI), respectively. Once the international community formulated a collective goal to fully characterize the whole human genome sequence, techniques for higher sequence throughput were developed to achieve this lofty objective.
Hitachi Laboratories (Tokyo, Japan) was among the first to answer this call, and in 1996 developed a high-throughput capillary array DNA sequencer later commercialized by Applied Biosystems. Amersham, plc (Amersham, U.K.) and Applied Biosystems brought to market fully automated sequencing using parallel analysis, thanks largely to major advances in miniaturization of robotic sample preparation, laboratory automation, and new enzymes and biochemicals. Working in parallel, the consortia sequenced the human genome in April 2003.
The Sanger method made possible the sequencing of the human genome and opened the door for innumerable applications in biology and medicine. Unfortunately, Sanger sequencing protocols were found to be insufficient to meet the throughput and automation needs of the growing application of genomic sequencing. The gels and polymers used as sieving separation media for fluorescently labeled DNA fragments, the insufficient capability to parallel process samples, and the automation difficulties catalyzed efforts to develop gel-less techniques capable of processing millions of samples in parallel.
Today, novel DNA sequencing platforms provide rapid, high-throughput, and relatively simple-to-use protocols. Projects that previously took multiple years using the Sanger method can now be completed in a matter of weeks. Instrumental to this drastic method acceleration is the acquisition of sequence data from amplified DNA fragments, rather than the timely, and costly, cloning of DNA fragments. Cost, however, remains a limiting factor in the adoption of NGS for many laboratories despite the cost-per-base being lower by multiple orders of magnitude when compared to the Sanger method.
Advantages of automated sample preparation for genomic sequencing
In the total testing process, the preanalytical phase is particularly vulnerable to errors. Mislabeling, contamination, or deviation from sample processing protocols can confound even the most elegant of analytical methods. In the case of genetic sequencing, preanalytical errors can be compounded with every thermocycle. DNA extraction protocols must be carefully optimized and rigorously followed. The largest bottleneck in the sample preparation process is library construction, requiring multiple time-consuming and error-prone steps. Between–sample, –day, and –technician variation is inevitable when considering the complexity of sequencing sample preparation. Often, locating the source of errors within the multiple steps and personnel required in manually prepared libraries can be impractical, if not impossible. Once the libraries are fully sized, PCR, normalization, and sequencing analytical steps follow, each involving careful calculation of reagent volumes and multistep procedures prone to human error.
When reproducibility and consistency are paramount, automation in the laboratory can be a great help. In addition to better process control, higher data quality, and accurate and efficient work flows, automated sample preparation for genomic sequencing can bring a level of repeatability between-run and between-technicians difficult to achieve by even the most experienced and diligent of laboratories.
It should be noted that human intervention cannot be fully eliminated. It is important to incorporate manual intervention between key steps in the process to ensure the robot is performing according to protocol. Internal controls run intermittently to check the instrument can also add confidence and head off systematic errors. As NGS is increasingly used for diagnostic purposes, precautions should be taken to validate runs with internal quality control derived quality reports. Also, depending on the application or sample size, automated sample processing may not make sense temporally or fiscally. There is generally a lengthy setup time, and automated methods tend to use larger reagent volumes than manual methods. Generally in the laboratory, there is a trade-off between speed and sensitivity, and high-throughput automation is not entirely immune to this relationship.
Table 1: Liquid handling robotics
There are many liquid handling robots on the market (Table 1), and choosing the right one to match your laboratory needs can be challenging. For NGS, selecting a system that is highly versatile and customizable can help ensure applicability across a range of sequencing applications. For example, the SPRIworks Systems (Beckman Coulter, Brea, CA) offers the ability to select your sample size (up to 96 per day) and the dry time. SPRIworks Systems also allow the researcher to choose when to perform the library construction, be it at end repair, A-tailing, or ligation steps. Unique to the SPRIworks platform, researchers can also set a size selection perimeter: none—150 bp +, small—150 bp to 350 bp, medium—250 bp to 450 bp, or large—350 bp to 700 bp. As pointed out by Dr. Shawn Levy during his presentation, “Automating High-Throughput Sample Prep for the Analysis of Complex Diseases Using Next-Gen Sequencing,” “… [the size selection parameter] is an exciting opportunity because now you’re not switching reagent or switching instrument types to have this size selection capability.” Finally, the SPRIworks System will take your setup inputs and precalculate the specific volumes of reagents needed to successfully extract and prepare the genetic material and run PCR. Dr. Shawn Levy summarizes his preference for the SPRIworks System as “a very convenient start to finish solution….”
Personal genomics and future innovation
A new generation of clinicians and research scientists are applying NGS beyond the original development goal and application of genomic sequencing. Following medical paradigms in personalized medicine, personal genomics with detailed analysis of individual genomic stretches is soon to be realized. The $1000 human genome sequence is the target with two companies, Life Technologies (Carlsbad, CA) and Illumina (San Diego, CA), leading the charge. While the clinical implications, such as early disease detection, companion diagnostics with personalized therapeutics, and improved patient outcomes are yet to be fully realized, the cheaper cost-per-base promises to usher sequencing research out of the large government-run laboratories, into core facilities, and on to individual laboratories. Market motivators likely to encourage future development of NGS automation include:
- Automated sample preparation can increase the efficiency of NGS clinical trials, thereby increasing the odds of success in a clinical development program
- When medicine has been proven safe and highly effective within a specific patient population delineated by rapid genome sequencing, its price can increase to reflect the value to the patient
- As effective treatment of disease eliminates the need for additional medication or treatments, cost savings can be applied to diagnostic costs
- Cost savings from fewer personnel can also be applied to minimize diagnostic costs.
Conclusion: The changing face of sequencing technology
Though not clear to what extent, NGS will surely play a significant role in the future of personalized medicine. There are many stakeholders (i.e., patients, clinicians, pathologists, laboratory technicians, pharmaceutical companies, and diagnostic test manufacturers), each with unique concerns and needs regarding the changing application of sequencing technologies. The Food and Drug Administration, partnering with the National Center for Biotechnology Information (NCBI), the National Institute of Standards and Technology (NIST), and the National Human Genome Research Institute (NHGRI), are currently in the process of developing evaluation protocols and standards in anticipation of NGS platforms being applied throughout biomedicine.1 To answer the growing demand for rapid, consistent genetic sequencing, liquid handling robotics offers a solution today. Offering a normalized work flow unattainable by manual means, automated sample preparation, amplification, and sequencing can increase DNA extraction yield and consistency and minimize between-technician bias.
- Ultra High Throughput Sequencing for Clinical Diagnostic Applications—Approaches to Assess Analytical Validity: Report from the Public Meeting (June 23, 2011). The Food and Drug Administration Web site: http://www.fda.gov/MedicalDevices/NewsEvents/WorkshopsConferences/ucm284442.htm. Accessed July 9, 2012.